WO2014084255A1 - 細胞観察装置、細胞観察方法及びそのプログラム - Google Patents
細胞観察装置、細胞観察方法及びそのプログラム Download PDFInfo
- Publication number
- WO2014084255A1 WO2014084255A1 PCT/JP2013/081894 JP2013081894W WO2014084255A1 WO 2014084255 A1 WO2014084255 A1 WO 2014084255A1 JP 2013081894 W JP2013081894 W JP 2013081894W WO 2014084255 A1 WO2014084255 A1 WO 2014084255A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- cell
- image
- pixel
- cells
- plastid
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 77
- 210000004027 cell Anatomy 0.000 claims abstract description 855
- 210000002706 plastid Anatomy 0.000 claims abstract description 102
- 238000000605 extraction Methods 0.000 claims description 103
- 239000000049 pigment Substances 0.000 claims description 72
- 238000003860 storage Methods 0.000 claims description 58
- 230000008569 process Effects 0.000 claims description 32
- 238000001514 detection method Methods 0.000 claims description 27
- 239000000284 extract Substances 0.000 claims description 27
- 239000000203 mixture Substances 0.000 claims description 16
- 239000006185 dispersion Substances 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 5
- 239000002131 composite material Substances 0.000 claims description 5
- 238000003786 synthesis reaction Methods 0.000 claims description 5
- 239000002356 single layer Substances 0.000 claims description 4
- 230000002194 synthesizing effect Effects 0.000 abstract description 4
- 238000009826 distribution Methods 0.000 abstract description 3
- JEBFVOLFMLUKLF-IFPLVEIFSA-N Astaxanthin Natural products CC(=C/C=C/C(=C/C=C/C1=C(C)C(=O)C(O)CC1(C)C)/C)C=CC=C(/C)C=CC=C(/C)C=CC2=C(C)C(=O)C(O)CC2(C)C JEBFVOLFMLUKLF-IFPLVEIFSA-N 0.000 description 83
- 235000013793 astaxanthin Nutrition 0.000 description 83
- 229940022405 astaxanthin Drugs 0.000 description 83
- 239000001168 astaxanthin Substances 0.000 description 83
- MQZIGYBFDRPAKN-ZWAPEEGVSA-N astaxanthin Chemical compound C([C@H](O)C(=O)C=1C)C(C)(C)C=1/C=C/C(/C)=C/C=C/C(/C)=C/C=C/C=C(C)C=CC=C(C)C=CC1=C(C)C(=O)[C@@H](O)CC1(C)C MQZIGYBFDRPAKN-ZWAPEEGVSA-N 0.000 description 82
- 229930002875 chlorophyll Natural products 0.000 description 80
- 235000019804 chlorophyll Nutrition 0.000 description 80
- ATNHDLDRLWWWCB-AENOIHSZSA-M chlorophyll a Chemical compound C1([C@@H](C(=O)OC)C(=O)C2=C3C)=C2N2C3=CC(C(CC)=C3C)=[N+]4C3=CC3=C(C=C)C(C)=C5N3[Mg-2]42[N+]2=C1[C@@H](CCC(=O)OC\C=C(/C)CCC[C@H](C)CCC[C@H](C)CCCC(C)C)[C@H](C)C2=C5 ATNHDLDRLWWWCB-AENOIHSZSA-M 0.000 description 80
- 210000004748 cultured cell Anatomy 0.000 description 59
- 238000010586 diagram Methods 0.000 description 45
- 230000006870 function Effects 0.000 description 29
- 238000012549 training Methods 0.000 description 27
- 238000012360 testing method Methods 0.000 description 22
- 239000000975 dye Substances 0.000 description 20
- 238000004519 manufacturing process Methods 0.000 description 19
- 230000035790 physiological processes and functions Effects 0.000 description 19
- 238000010191 image analysis Methods 0.000 description 18
- 239000002245 particle Substances 0.000 description 18
- 238000011534 incubation Methods 0.000 description 17
- 238000012545 processing Methods 0.000 description 16
- 241000168525 Haematococcus Species 0.000 description 15
- 230000012010 growth Effects 0.000 description 15
- 238000003384 imaging method Methods 0.000 description 15
- 230000008859 change Effects 0.000 description 14
- 238000007637 random forest analysis Methods 0.000 description 14
- 230000005484 gravity Effects 0.000 description 13
- 239000002609 medium Substances 0.000 description 13
- 241000894007 species Species 0.000 description 13
- 238000004458 analytical method Methods 0.000 description 12
- 238000005259 measurement Methods 0.000 description 12
- 241000168517 Haematococcus lacustris Species 0.000 description 11
- 239000000126 substance Substances 0.000 description 11
- 238000004364 calculation method Methods 0.000 description 10
- 238000011109 contamination Methods 0.000 description 10
- 210000003495 flagella Anatomy 0.000 description 10
- 239000001963 growth medium Substances 0.000 description 10
- IAZDPXIOMUYVGZ-UHFFFAOYSA-N Dimethylsulphoxide Chemical compound CS(C)=O IAZDPXIOMUYVGZ-UHFFFAOYSA-N 0.000 description 9
- 235000021466 carotenoid Nutrition 0.000 description 9
- 150000001747 carotenoids Chemical class 0.000 description 9
- 230000002596 correlated effect Effects 0.000 description 9
- 238000002835 absorbance Methods 0.000 description 8
- 238000004113 cell culture Methods 0.000 description 8
- 238000000611 regression analysis Methods 0.000 description 8
- 241000195493 Cryptophyta Species 0.000 description 6
- 241000233866 Fungi Species 0.000 description 6
- KWYUFKZDYYNOTN-UHFFFAOYSA-M Potassium hydroxide Chemical compound [OH-].[K+] KWYUFKZDYYNOTN-UHFFFAOYSA-M 0.000 description 6
- 238000000855 fermentation Methods 0.000 description 6
- 230000004151 fermentation Effects 0.000 description 6
- 239000012737 fresh medium Substances 0.000 description 6
- 238000010801 machine learning Methods 0.000 description 6
- 230000004044 response Effects 0.000 description 6
- 240000004808 Saccharomyces cerevisiae Species 0.000 description 5
- 230000010261 cell growth Effects 0.000 description 5
- 239000008188 pellet Substances 0.000 description 5
- 239000006228 supernatant Substances 0.000 description 5
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 4
- 241000195649 Chlorella <Chlorellales> Species 0.000 description 4
- MBMBGCFOFBJSGT-KUBAVDMBSA-N all-cis-docosa-4,7,10,13,16,19-hexaenoic acid Chemical compound CC\C=C/C\C=C/C\C=C/C\C=C/C\C=C/C\C=C/CCC(O)=O MBMBGCFOFBJSGT-KUBAVDMBSA-N 0.000 description 4
- 238000005119 centrifugation Methods 0.000 description 4
- 238000000513 principal component analysis Methods 0.000 description 4
- 230000002123 temporal effect Effects 0.000 description 4
- 238000012935 Averaging Methods 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 3
- 239000006143 cell culture medium Substances 0.000 description 3
- 238000012258 culturing Methods 0.000 description 3
- 210000003298 dental enamel Anatomy 0.000 description 3
- 230000008034 disappearance Effects 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 239000011521 glass Substances 0.000 description 3
- 238000003709 image segmentation Methods 0.000 description 3
- 238000011081 inoculation Methods 0.000 description 3
- 230000003834 intracellular effect Effects 0.000 description 3
- 239000010410 layer Substances 0.000 description 3
- 239000007788 liquid Substances 0.000 description 3
- 238000001000 micrograph Methods 0.000 description 3
- 238000002360 preparation method Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 238000010186 staining Methods 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 210000005253 yeast cell Anatomy 0.000 description 3
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 238000012313 Kruskal-Wallis test Methods 0.000 description 2
- 238000011481 absorbance measurement Methods 0.000 description 2
- 235000013334 alcoholic beverage Nutrition 0.000 description 2
- JAZBEHYOTPTENJ-JLNKQSITSA-N all-cis-5,8,11,14,17-icosapentaenoic acid Chemical compound CC\C=C/C\C=C/C\C=C/C\C=C/C\C=C/CCCC(O)=O JAZBEHYOTPTENJ-JLNKQSITSA-N 0.000 description 2
- 238000009395 breeding Methods 0.000 description 2
- 230000001488 breeding effect Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 239000000470 constituent Substances 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 2
- 235000020669 docosahexaenoic acid Nutrition 0.000 description 2
- 229940090949 docosahexaenoic acid Drugs 0.000 description 2
- 235000020673 eicosapentaenoic acid Nutrition 0.000 description 2
- 229960005135 eicosapentaenoic acid Drugs 0.000 description 2
- JAZBEHYOTPTENJ-UHFFFAOYSA-N eicosapentaenoic acid Natural products CCC=CCC=CCC=CCC=CCC=CCCCC(O)=O JAZBEHYOTPTENJ-UHFFFAOYSA-N 0.000 description 2
- 230000031700 light absorption Effects 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 230000004660 morphological change Effects 0.000 description 2
- 229910052757 nitrogen Inorganic materials 0.000 description 2
- 230000003071 parasitic effect Effects 0.000 description 2
- 230000001376 precipitating effect Effects 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 238000010298 pulverizing process Methods 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 230000036962 time dependent Effects 0.000 description 2
- 238000003260 vortexing Methods 0.000 description 2
- FZIPCQLKPTZZIM-UHFFFAOYSA-N 2-oxidanylpropane-1,2,3-tricarboxylic acid Chemical compound OC(=O)CC(O)(C(O)=O)CC(O)=O.OC(=O)CC(O)(C(O)=O)CC(O)=O FZIPCQLKPTZZIM-UHFFFAOYSA-N 0.000 description 1
- RBTBFTRPCNLSDE-UHFFFAOYSA-N 3,7-bis(dimethylamino)phenothiazin-5-ium Chemical compound C1=CC(N(C)C)=CC2=[S+]C3=CC(N(C)C)=CC=C3N=C21 RBTBFTRPCNLSDE-UHFFFAOYSA-N 0.000 description 1
- 102000007469 Actins Human genes 0.000 description 1
- 108010085238 Actins Proteins 0.000 description 1
- 241000894006 Bacteria Species 0.000 description 1
- 241001536324 Botryococcus Species 0.000 description 1
- 229920002101 Chitin Polymers 0.000 description 1
- KRKNYBCHXYNGOX-UHFFFAOYSA-K Citrate Chemical compound [O-]C(=O)CC(O)(CC([O-])=O)C([O-])=O KRKNYBCHXYNGOX-UHFFFAOYSA-K 0.000 description 1
- KCXVZYZYPLLWCC-UHFFFAOYSA-N EDTA Chemical compound OC(=O)CN(CC(O)=O)CCN(CC(O)=O)CC(O)=O KCXVZYZYPLLWCC-UHFFFAOYSA-N 0.000 description 1
- VTLYFUHAOXGGBS-UHFFFAOYSA-N Fe3+ Chemical compound [Fe+3] VTLYFUHAOXGGBS-UHFFFAOYSA-N 0.000 description 1
- 241000235343 Saccharomycetales Species 0.000 description 1
- 229920002472 Starch Polymers 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000001467 acupuncture Methods 0.000 description 1
- 239000003963 antioxidant agent Substances 0.000 description 1
- 230000003078 antioxidant effect Effects 0.000 description 1
- 235000006708 antioxidants Nutrition 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 150000001514 astaxanthins Chemical class 0.000 description 1
- 235000013405 beer Nutrition 0.000 description 1
- 230000003570 biosynthesizing effect Effects 0.000 description 1
- 230000001851 biosynthetic effect Effects 0.000 description 1
- 210000000746 body region Anatomy 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 229910002092 carbon dioxide Inorganic materials 0.000 description 1
- 230000032823 cell division Effects 0.000 description 1
- 210000000170 cell membrane Anatomy 0.000 description 1
- 210000002421 cell wall Anatomy 0.000 description 1
- 230000002759 chromosomal effect Effects 0.000 description 1
- 239000006059 cover glass Substances 0.000 description 1
- 210000004292 cytoskeleton Anatomy 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000010432 diamond Substances 0.000 description 1
- YJHDFAAFYNRKQE-YHPRVSEPSA-L disodium;5-[[4-anilino-6-[bis(2-hydroxyethyl)amino]-1,3,5-triazin-2-yl]amino]-2-[(e)-2-[4-[[4-anilino-6-[bis(2-hydroxyethyl)amino]-1,3,5-triazin-2-yl]amino]-2-sulfonatophenyl]ethenyl]benzenesulfonate Chemical compound [Na+].[Na+].N=1C(NC=2C=C(C(\C=C\C=3C(=CC(NC=4N=C(N=C(NC=5C=CC=CC=5)N=4)N(CCO)CCO)=CC=3)S([O-])(=O)=O)=CC=2)S([O-])(=O)=O)=NC(N(CCO)CCO)=NC=1NC1=CC=CC=C1 YJHDFAAFYNRKQE-YHPRVSEPSA-L 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 108010049975 fluorescein isothiocyanate-wheat germ agglutinin Proteins 0.000 description 1
- 239000007850 fluorescent dye Substances 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 230000002538 fungal effect Effects 0.000 description 1
- 239000001056 green pigment Substances 0.000 description 1
- 235000013402 health food Nutrition 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 238000011177 media preparation Methods 0.000 description 1
- 239000013028 medium composition Substances 0.000 description 1
- 229960000907 methylthioninium chloride Drugs 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000004899 motility Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000029553 photosynthesis Effects 0.000 description 1
- 238000010672 photosynthesis Methods 0.000 description 1
- 230000000243 photosynthetic effect Effects 0.000 description 1
- 230000004962 physiological condition Effects 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 239000001054 red pigment Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 235000020083 shōchū Nutrition 0.000 description 1
- 239000008107 starch Substances 0.000 description 1
- 235000019698 starch Nutrition 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 235000021122 unsaturated fatty acids Nutrition 0.000 description 1
- 150000004670 unsaturated fatty acids Chemical class 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/695—Preprocessing, e.g. image segmentation
-
- C—CHEMISTRY; METALLURGY
- C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
- C12M—APPARATUS FOR ENZYMOLOGY OR MICROBIOLOGY; APPARATUS FOR CULTURING MICROORGANISMS FOR PRODUCING BIOMASS, FOR GROWING CELLS OR FOR OBTAINING FERMENTATION OR METABOLIC PRODUCTS, i.e. BIOREACTORS OR FERMENTERS
- C12M1/00—Apparatus for enzymology or microbiology
- C12M1/34—Measuring or testing with condition measuring or sensing means, e.g. colony counters
-
- C—CHEMISTRY; METALLURGY
- C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
- C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
- C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
- C12Q1/02—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving viable microorganisms
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
- G01N15/14—Optical investigation techniques, e.g. flow cytometry
- G01N15/1429—Signal processing
- G01N15/1433—Signal processing using image recognition
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/60—Editing figures and text; Combining figures or text
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
- G06T7/0014—Biomedical image inspection using an image reference approach
- G06T7/0016—Biomedical image inspection using an image reference approach involving temporal comparison
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/698—Matching; Classification
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/10—Investigating individual particles
- G01N2015/1006—Investigating individual particles for cytology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
Definitions
- the present invention relates to a cell observation apparatus, a cell observation method, and a program for observing the state of a cell using a microscopic image of the cell.
- fermentation can be performed using single-cell yeast to produce various alcoholic beverages such as beer and shochu.
- it is performed to determine the physiological state of yeast used for such fermentation before fermentation and to predict the influence on subsequent fermentation (see, for example, Patent Document 2).
- it grasping the physiological state of yeast cells to be used for production in advance predicts the success or failure of fermentation, leading to high quality and stable products. Is necessary to get.
- Microalgae mainly refers to single-cell photosynthetic organisms. This microalgae converts light energy into chemical energy by photosynthesis, and uses the converted energy for its own survival and growth. In addition, depending on the type of microalgae, there are those that biosynthesize hydrocarbons and essential unsaturated fatty acids (for example, DHA (Docosa Hexaenoic Acid), EPA (Eicosa Pentaenoic Acid), etc.), starch, pigments, etc. as useful components. Industrial utilization of biosynthetic functions is expected.
- DHA Docosa Hexaenoic Acid
- EPA Ecosa Pentaenoic Acid
- microalgae cells It is important to properly grasp the physiological state of the cells of the microalgae when trying to produce the above useful components with high efficiency using the microalgae. That is, the physiological state of microalgae cells varies greatly depending on the growth conditions of the surrounding environment, such as medium composition, carbon dioxide concentration, light intensity, culture temperature, and cell density. And microalgae cells also change the amount of production and accumulation of useful components according to their physiological state.
- the physiology of the cells of microalgae in culture It is necessary to grasp the state and grasp the production amount of useful ingredients.
- contamination of other species in the medium in which the cells are cultured may affect the physiological state of the microalgae cells. In many cases, the influence on the physiological state due to the contamination of other species becomes a problem in the production process of useful components by cells. Therefore, detection of other species mixed in the culture medium described above is also important for highly efficient production of useful components using microalgae cells.
- Haematococcus pluvialis which is a kind of microalgae
- This Haematococcus pluvialis is highly useful industrially by biosynthesizing the red antioxidant astaxanthin, which is also used as a health food.
- Haematococcus pluvialis shows various cell morphology reflecting the physiological state of the cell.
- the amount of astaxanthin accumulated also varies depending on the conditions during culture (see, for example, Non-Patent Document 1).
- a plurality of types of culture strains identified as Haematococcus pluvialis were collected from all over the world, and contamination of organisms other than Haematococcus pluvialis was examined. As a result, surprisingly, contamination was confirmed in all cultured strains including strains used industrially. Therefore, grasping the physiological state of cells of Haematococcus pluvialis, the accumulated amount of useful component astaxanthin, and the contamination rate of other species are important issues in industrial use.
- Patent Document 3 discloses a method for evaluating the physiological state of yeast using a quantitative value of cell morphology.
- Patent Document 1 quantifies the amount of Haematococcus pluvialis astaxanthin by extracting plastids from cells with dimethyl sulfoxide and measuring absorbance at 492 nm and 750 nm. The method is shown. However, the measurement requires extraction of plastids from a large number of microalgal cells, and it takes time to extract the plastids.
- Non-Patent Document 4 a method of specifically staining a fungus that is a fungus parasitizing diatom cells that are microalgae with a calcofluor white that binds to chitin that is a component of the cell wall of the fungus Is shown (Non-Patent Document 4).
- Non-Patent Document 5 Parasodyderma sedebokerensis zoosporangium parasitic on Haematococcus pluviaris was stained with FITC-WGA (Non-patent Document 5).
- Patent Document 2 and Non-Patent Document 5 require a fluorescence microscope, and thus are not suitable for discrimination in the field and production sites.
- the growth status of microalgae cells, the status of contamination of other species (eg, parasitic bacteria) mixed in the culture medium of microalgae, and the production of useful substances by the cells of microalgae There is no easy way to monitor in real time.
- the present invention has been made in view of such circumstances, and in order to improve the production amount of useful substances by cells such as microalgae, the development of culture conditions and the breeding of high production substances of useful substances are performed. It is an object of the present invention to provide a cell observation apparatus, a cell observation method, and a program for monitoring in real time each of the contamination status of other species mixed in the culture medium and the production amount of useful substances by cells of microalgae.
- the cell observation apparatus is configured to detect pixels at the edge of an image of a cell in the captured image from a captured image obtained by capturing a single layer of cells (a state in which cells are arranged on only one layer without overlapping on a plane).
- a contour extraction unit that extracts and generates an edge image composed of the extracted edge pixels, and extracts a chromophoric pixel of a cell image in the captured image, and generates a chromophoric image composed of the extracted chromophoric pixels.
- the synthesized image obtained by superimposing the edge image and the chromophoric image the cell image area and the background image area in the captured image are represented by the luminance value of the pixel.
- an image composition unit for detecting an image area of a cell in the captured image.
- the cell observation apparatus of the present invention detects an image area of a cell in which the plastid is present in the composite image as a target cell image, detects an image area of a cell in which no plastid is present as a non-target cell image, and It is further characterized by further comprising a cell shape detection unit for obtaining a ratio of the non-target cell image in the cell image.
- the cell observation apparatus of the present invention is further characterized by further comprising a pigment value extraction unit that calculates the amount of plastids, which is the amount of the plastids, from the luminance value of the plastids in the image area of the cells.
- the cell observation apparatus of the present invention obtains an average luminance value of a plastid from the image area of the cell, extracts a plastid from the cell obtained by imaging the captured image, obtains an amount of the plastid, A regression equation with the amount of plastids per cell is obtained in advance and stored in the storage unit, and the pigment value extraction unit obtains an average luminance value in the cell image, and obtains the amount of plastids per cell from the regression equation. It is characterized by that.
- the contour extraction unit extracts edge pixels of the cell image in the captured image from the captured image obtained by capturing one layer of cells, and generates an edge image including the extracted edge pixels.
- a chromophoric region extracting unit that extracts a chromophoric pixel of a cell image in the captured image and generates a plastid image composed of the extracted chromophoric pixels;
- the image synthesizing unit converts the cell image area and the background image area in the captured image by dispersion of luminance values of the pixels. And an image composition process for detecting an image area of a cell in the captured image.
- the program of the present invention is a program for causing a computer to execute an operation of a cell observation device for observing the shape of a cell.
- the computer uses a captured image obtained by imaging a single cell to detect the edge of an image of a cell in the captured image.
- Contour extraction means for extracting pixels and generating edge images composed of the extracted edge pixels, extracting chromophore pixels of the cell image in the captured image, and extracting the chromophoric image composed of the extracted chromophore pixels
- the chromophoric region extracting means to generate, the cell image region and the background image region in the captured image, the luminance value of the pixel
- the edge image extracted from the captured image that is a microphotograph of microalgae and the plastid image are superimposed, the image of the cell in the captured image can be extracted with higher accuracy than in the past, and the cell
- the overall shape of the cells and the proportion of plastid regions (for example, regions of the produced product) in the cells can be easily detected.
- an indicator of the physiological state of the cell, the contamination status of other species, and the accumulation status of the colored plastids are obtained as quantitative values, and the physiological status of the cell and the production status of useful components can be obtained. Easy to grasp.
- FIG. 1 It is a schematic block diagram which shows the structural example of the cell observation apparatus 1 by one Embodiment of this invention. An image displayed by superimposing an analysis result on a captured image (bright field image) captured by the imaging apparatus 100 is shown. It is a figure which shows the kind of numerical information regarding each cell written and memorize
- 2 is a graph showing changes in the amount of astaxanthin and chlorophyll accumulated measured by the cell observation apparatus 1 after inoculating Haematococcus on a fresh medium, and a representative cell image.
- the amount of pigment ( ⁇ g / ml) obtained by converting the amount of pigment per cell of chlorophyll accumulated in the observed cells (hematococcus) measured per 1 ml of the culture solution, and R pixel, G pixel, and pixel in the cell image It is a graph which shows the correlation with the pigment
- the amount of dye ( ⁇ g / ml) obtained by converting the amount of dye per cell of astaxanthin accumulated in the actually observed cells to be observed (haematococcus) ( ⁇ g / ml), the R pixel, the G pixel, and the pixel in the cell image It is a graph which shows the correlation with the pigment amount of astaxanthin estimated from the multiple regression equation using the average value of the luminance value of B pixel.
- the amount of dye ( ⁇ g / ml) obtained by converting the amount of dye per cell of astaxanthin accumulated in the actually observed cells to be observed (haematococcus) ( ⁇ g / ml), the R pixel, the G pixel, and the pixel in the cell image
- Cultured time AreaRatio indicating the ratio between the area and the area of the cell plastids in cells of the cell image: a diagram showing the correspondence between the (unit ⁇ m 2 / ⁇ m 2). It is a figure which shows the correspondence with InnerTotalBlueIntensity (unit: color (unit)) which is a numerical value which shows the addition value of the luminance value of the blue channel (B pixel) of the plastid body in the cell of the cell image of a cultured cell and the cell image of a cultured cell. It is a figure which shows the correspondence of culture
- OuterRoundFitness is a diagram showing the correspondence between the (unit ⁇ m 2 / ⁇ m 2). It is a figure which shows the correspondence of InnerMeanRedIntensity (unit: color
- FIG. 46 is a table showing the correlation coefficient between each of the discriminant function LD1 and discriminant function LD2 and the parameters shown in FIGS. It is a figure which shows a response
- FIG. 1 is a schematic block diagram showing a configuration example of a cell observation device 1 according to an embodiment of the present invention.
- a single-cell organism cell is imaged by a color camera attached to a microscope, and is an image that is a color bright-field image composed of pixels composed of RGB (Red, Green, Blue) pixels. From the image, the physiological state of the cell in the growth process and the mixed other species mixed in the culture medium for culturing the cell are analyzed.
- RGB Red, Green, Blue
- the cell observation device 1 includes a control unit 11, a hue adjustment unit 12, a contour extraction unit 13, a plastid region extraction unit 14, an image division unit 15, an image synthesis unit 16, and a cell region extraction unit 17.
- cells to be observed for physiological conditions are exemplified by cells of “Haematococcus pluvialis” (hereinafter referred to as “hematococcus”), which are single-celled organisms of microalgae. The image of the species will be described as a key fungus. Hematococcus initially accumulates chlorophyll as a plastid (green pigment) as a physiological state, but as it grows, it accumulates a plastid of a useful substance called astaxanthin (red pigment).
- the imaging apparatus 100 is provided with a CCD camera in a microscope.
- the imaging apparatus 100 captures hematococcus cells cultured in the culture medium 300 in the culture vessel 200 at a predetermined magnification, and captures the captured image with respect to the cell observation apparatus 1. Output.
- the observer selects a region in which cells are arranged in a single layer without contacting other cells as an imaging region for capturing a captured image, and images the hematococcus cells in the culture medium 300 by the imaging device 100. .
- FIG. 2 shows an image displayed by superimposing the analysis result on a captured image (bright field image) captured by the imaging apparatus 100.
- the cell image is composed of an outer periphery and a plastid region.
- the image in FIG. 2 is an image displayed on the display unit 24.
- the observer can arbitrarily display the processing result of each unit on the display unit 24.
- the processed image of the captured image is written and stored in the storage unit 22 by each unit that has performed each process.
- the ratio of the major axis to the minor axis is obtained by dividing the area of the cell indicated by the pixel value and the maximum width (OuterLengthAxisLength) at the outermost edge of the cell image by the minimum width (OuterShortAxisLength).
- the ratio L / S) is described in each cell image.
- an identification number (ID) for identifying each cell image is described at the center of the cell image.
- the control unit 11 writes and stores data of a captured image (bright field image) obtained by imaging the cells of the culture medium 300 in the culture container 200 obtained from the imaging device 100 in the image storage unit 21.
- the hue adjustment unit 12 reads the captured image from the image storage unit, adjusts the RGB luminance value (gradation) of each pixel of the captured image, and makes the background gray based on the luminance distribution in the read captured image. Adjust.
- the contour extraction unit 13 extracts an edge of an image in a captured image whose hue is adjusted by the Canny method and generates an edge image.
- the Canny method is described in “Canny, J., A computational approach to edge detection, IEEE Trans. Pattern Analysis and Machine Intelligence, 8: 679-714, 1986”.
- the plastid region extraction unit 14 binarizes a pixel region (a plastid region in a cell to be observed, which will be described later) having a similar luminance value in the captured image in the captured image by the Otsu method. Extract by Here, regarding the Otsu method, “Otsu N, A threshold selection method from gray-level”. histograms, IEEE Transactions on Systems, Man and Cybernetics, 9 (1): 62-66, 1979 ".
- the image dividing unit 15 superimposes the edge image extracted by the contour extracting unit 13 and deletes the edge that overlaps the pixel region, thereby obtaining a new edge image.
- the image segmentation unit 15 uses the edge information of the edge image to extract a segmented image (region segmentation) process of the captured image to extract a region of the cell image in the captured image that is an object to be extracted. Perform by law.
- a segmented image region segmentation
- the image composition unit 16 divides the divided segments into cell segments whose luminance value variance is larger than a predetermined threshold value and background segments whose luminance value variance is smaller than a predetermined threshold value. Select as the cell segment corresponding to the imaged area of the cell.
- the image synthesis unit 16 synthesizes each of the cell segment and the background segment, generates a cell image and a background image, respectively, and generates a synthesized image.
- the cell region extraction unit 17 obtains the similarity between the cell image region and the circular model image by the chordogram method, and sets a region that is equal to or higher than the predetermined similarity as a round region and a region that is less than the similarity. . Further, the cell region extraction unit 17 integrates regions that are not round.
- the calculation of similarity to a circular model image using the method of the chordogram is “Toshev, A., Taskar, B., Danilidis, K., Object detection via boundary structure segmentation, Computer -957, 2010 ".
- the cell structure extraction unit 18 obtains numerical information (described later) for individual cells and a plurality of cells based on the cell outer shape of the cell image in the captured image and the cell constituent parts (plastid region and non-plastid region) in the cell image.
- the obtained numerical information is written and stored in the table storage unit 23 for each cell image and each captured image from which the cell image is extracted.
- the cell shape detection unit 19 determines whether or not the cell is an observation target cell based on the presence or absence of the obtained plastid region of the cell constituent unit. That is, the cell shape detection unit 19 determines whether the cell indicated by the cell image is a hematococcus of the observation target cell or a mixed mold of other species based on the presence or absence of the plastid region. At this time, the cell shape detection unit 19 determines that the cell image with the plastid region is hematococcus of the observation target cell, and determines that the cell image without the plastid region is the acupuncture fungus of the non-observation target cell. In addition, the cell morphology detection unit 19 estimates the cell growth process from the cell outer shape of the cell (the ratio between the major axis and the minor axis of the cell image described later).
- the pigment value extraction unit 20 obtains an average value of luminance values of each of the RGB pixels in the cell image, and uses a regression equation that is obtained in advance using the average value and stored in the storage unit 22. The amount of accumulated pigment in the plastid region is estimated.
- the amount of astaxanthin and chlorophyll are estimated.
- FIG. 3 is a diagram showing the types of numerical information relating to individual cells written and stored in the table storage unit 23.
- Each type of data of this numerical information Name, ID, Type, OuterArea, OuterOutlineLength, OuterCenterX, OuterCenterY, OuterMaxRadius, OuterLongAxisLength, OuterShortAxisLength, OuterAxisRatio (L / S), Round fitness, Chordiogram distance, OuterRedIntensity, OuterGreenIntensity, OuterBlueIntensity, InnerArea , InnerOutlineLength, InnerRedIntensity, InnerGreenIntensity, InnerBlueIntens It contains ty.
- each of the numerical information in FIG. 3 will be described.
- Name indicates the file name of the image data of the cell image.
- the ID is an identification number indicating a region of the cell image recognized as a cell, and may include position information (coordinate values) in the captured image.
- Type indicates the type of cell indicated by the cell image with the ID attached.
- the target observation target microalgae observation target cells for example, cell images with IDs of 8 and 10, etc. in FIG. 2
- contamination that is an organism other than the observation target cells in the captured image
- OuterArea indicates the area of the cell image, and the area is set by the number of pixels.
- OuterOutlineLength is the length of the outer periphery of the cell image, and is a numerical value represented by the number of pixels arranged on the outer periphery of the cell image.
- OuterCenterX is a numerical value indicating the x-coordinate (pixel value, that is, the number of pixels from the left end of the captured image) of the center of the cell image in the xy coordinate system of the captured image.
- OuterCenterY is a numerical value indicating the y-coordinate (pixel value, that is, the number of pixels from the upper end of the captured image) of the center of the cell image in the xy coordinate system of the captured image.
- OuterMaxRadius is a numerical value indicating the maximum value (maximum width) of the distance from the coordinates of the center of gravity to the outer periphery of the captured image in terms of the number of pixels.
- Outer Long Axis Length (L) is a numerical value indicating the dimension of the maximum width portion of the cell image, that is, the length of the long axis in the number of pixels.
- Outer Short Axis Length (S) is a numerical value indicating the dimension of the minimum width portion of the cell image, that is, the length of the short axis in the number of pixels.
- OuterAxisRatio (L / S) is a numerical value indicating the ratio L / S between OuterLongAxisLength and OuterShortAxisLength.
- Round fitness is a numerical value indicating the degree of conformity of the shape of the cell image to the circle. Also, Round fitness is calculated from the equation “4 ⁇ ⁇ area ⁇ square of perimeter”.
- the chordogram distance is a numerical value indicating the similarity between the circle model image and the comparison result of the target image by the chordogram, and obtains the similarity indicating the degree of similarity to the model image close to a circle.
- OuterRedIntensity is a numerical value indicating an average value of luminance values of the red channel (R pixel) in the cell.
- OuterGreenIntensity is a numerical value indicating an average value of luminance values of the green channel (G pixel) in the cell.
- OuterBlueIntensity is a numerical value indicating the average value of the luminance values of the blue channel (B pixel) in the cell.
- InnerArea is a numerical value indicating the area of the plastid region in terms of the number of pixels in the region.
- InnerOutlineLength is a numerical value indicating the outer periphery of the pigment body region in terms of the number of pixels.
- InnerRedIntensity is the average luminance value of the red channel (R pixel) in the chromophore region.
- InnerGreenIntensity is the average luminance value of the green channel in the chromophore region.
- InnerBlueIntensity is the average luminance value of the blue channel in the chromophore region.
- the cell structure extraction unit 18 counts the number of pixels in the cell image, and writes the number of pixels in the table storage unit 23 as the cell area OuterArea of the numerical information in FIG.
- the cell structure extraction unit 18 counts the number of pixels arranged on the outer periphery of the cell image, and writes the number of pixels on the outer periphery into the table storage unit 23 as the cell outer length OuterLength shown in FIG. .
- the cell structure extraction unit 18 obtains the center of gravity from the area of the cell image, obtains the x-coordinate and y-coordinate of the center of gravity in the captured image, and writes and stores them in the table storage unit 23 as OuterCenterX and OuterCenterY, respectively.
- the x coordinate is the number of pixels from the left end of the captured image
- the y coordinate is the number of pixels from the upper end of the captured image.
- the cell structure extraction unit 18 counts the pixel value having the maximum width from the coordinates of the center of gravity (OuterCenterX, OuterCenterY) to the outer periphery of the cell image, and writes the counted pixel value as OuterMaxRadius in the numerical information of FIG. To remember.
- the cell structure extraction unit 18 counts the number of pixels having the length of the major axis (major axis) of the cell image, and writes this number of pixels as the Outer Long Axis Length (L) in the numerical information of FIG. 3 in the table storage unit 23 for storage.
- the cell structure extraction unit 18 counts the number of pixels of the length of the short axis (minor axis) of the cell image, and writes this number of pixels in the table storage unit 23 as Outer Short Axis Length (S) in the numerical information of FIG.
- S Outer Short Axis Length
- the cell structure extraction unit 18 divides OuterLongAxisLength by OuterShortAxisLength, and writes and stores the division result in the table storage unit 23 as OuterAxisRatio (L / S) in the numerical information of FIG.
- the cell structure extraction unit 18 calculates “(4 ⁇ ⁇ OuterArea) / (OuterOutlineLength) 2 ”, calculates the fitness to the circle, and stores this fitness as round fitness in the numerical information of FIG.
- the data is written in the unit 23 and stored.
- the cell structure extraction unit 18 calculates the distance and angle between any two points and the angle of the perpendicular to the outer periphery of these two points in the circular model image and the cell image, and calculates the difference between these histograms (Chordiogram). The difference is obtained as the similarity to the model image, and is written and stored in the table storage unit 23 as the chordogram distance in the numerical information of FIG.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the R pixels in the cell image, and writes this average value in the table storage unit 23 as OuterRedIntensity in the numerical information of FIG.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the G pixels in the cell image, and writes this average value in the table storage unit 23 as OuterGreenIntensity in the numerical information of FIG.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the B pixels in the cell image, and writes the average value in the table storage unit 23 as OuterBlueIntensity in the numerical information of FIG.
- the cell structure extraction unit 18 obtains the area of the plastid region extracted by the plastid region extraction unit 14 as a count value obtained by counting the number of pixels included in the plastid region, and the count value is the numerical value of FIG. As the Inner Area in the information, it is written and stored in the table storage unit 23.
- the cell structure extraction unit 18 counts the number of pixels arranged on the outer periphery of the plastid region, and writes the number of pixels on the outer periphery to the table storage unit 23 as the inner outer length of the plastid region shown in FIG. To remember.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the R pixels in the plastid region, and writes this average value in the table storage unit 23 as InnerRedIntensity in the numerical information of FIG.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the G pixels in the plastid region, and writes this average value in the table storage unit 23 as InnerGreenIntensity in the numerical information of FIG.
- the cell structure extraction unit 18 calculates the average value of the luminance values of the B pixels in the plastid region, and writes this average value in the table storage unit 23 as InnerBlueIntensity in the numerical information of FIG.
- FIG. 4 is a diagram illustrating types of numerical information related to cells in a plurality of captured images that are written and stored in the table storage unit 23.
- Each type of data of the numerical information includes Name, Touch, Continant, Algae, Others, Cell count, Astaxanthin predicator, Astaxanthin, Chlorophyllic predictor, and Chlorophyll.
- Each of the numerical information in FIG. 4 will be described.
- “Name” indicates a folder name of image data of a captured image composed of a plurality of image files of cell images.
- Touch is a numerical value indicating the number of images of cells in contact with the edge of the captured image.
- Algae is a numerical value indicating the number of target cells (in this embodiment, single cells of microalgae) that are observation targets in the captured image. This target cell is a cell having a similarity with a circular model image of less than 0.5 and a plastid region.
- “Continant” is a numerical value indicating the number of non-target cells (in this embodiment, for example, a single mold of mold) in the captured image.
- Non-target cells not intended for this purpose are cells having a similarity with a circular model image of less than 0.5 and having no plastid region. Others shows a non-cell image that is neither a target cell nor a non-target cell. This non-cell indicates a non-determinable cell having a similarity with the model image of 0.5 or more.
- the similarity for determining the target cell, the non-target cell, and the non-cell is obtained by the method of the chordogram.
- Cell count indicates the total number of cells detected in the captured image, that is, the total number of cells in contact with the edge of the captured image, the number of target cells, and the number of non-target cells. In this embodiment, it is a numerical value obtained by adding the number of cells in contact with the edge of the captured image, the number of microalgae, the number of amber mold and the number of non-cells.
- Astaxanthin predictor is a numerical value used for measuring the amount of astaxanthin (in this embodiment, a numerical value obtained by dividing the average value of the luminance value of the R pixel by the average value of the luminance value of the B pixel).
- the average value of the luminance values of the R pixels and the average value of the luminance values of the B pixels are average values obtained by averaging the average values in the target cells included in the captured image in all the target cells included in the captured image.
- the average value in the target cell is an average of the luminance values of the pixels in all the pixels included in the target cell.
- Astaxanthin indicates the amount of astaxanthin (regression formula: 137.9 ⁇ predictor-174.3) obtained from a regression curve described later per pixel using Astaxanthin predictor, which is an average value of luminance values of R pixels.
- Chlorophyll predicator is a numerical value used for measuring the chlorophyll amount (in this embodiment, a numerical value obtained by dividing an average value of luminance values of G pixels by an average value of luminance values of B pixels).
- the average value of the luminance values of the G pixels and the average value of the luminance values of the B pixels are average values obtained by averaging the average values in the target cells included in the captured image in all the target cells included in the captured image.
- the average value in the target cell is an average of the luminance values of the pixels in all the pixels included in the target cell.
- Chlorophyll indicates the amount of chlorophyll (regression equation: 284.7 ⁇ predictor-369.1) obtained from a regression curve described later per pixel, using Chlorophyll predicator, which is the average value of luminance values of G pixels.
- the cell structure extraction unit 18 reads the type of the cell image in the table storage unit 23, counts the number of each cell, and obtains the calculated count value. 23 is written and stored as numerical information shown in FIG. Further, for the cell count, the sum of the respective cell numbers is obtained, and the obtained calculated value is written and stored as numerical information shown in FIG.
- FIG. 5 shows the luminance which is the ratio of the pigment amount per cell of astaxanthin accumulated in the observation target cell (hematococcus) and the average value of the luminance values of the R pixel and the B pixel in the pixel in the cell image.
- the vertical axis represents the accumulated amount of astaxanthin per cell (pg: picogram), and the horizontal axis represents the average value of the luminance value of the R pixel of the pixel in the cell image divided by the average value of the B pixel.
- the luminance value ratio is shown.
- the regression equation for calculating astaxanthin is obtained in advance as follows, and is written and stored in the storage unit 22 in advance. Perform the above-described image analysis of the captured image to obtain numerical information in the captured image described above (other than the accumulated amount of pigment), and the average value of the luminance value of the R pixel and the average value of the luminance value of the B pixel of the cell image The luminance ratio is obtained.
- dye was extracted from the cell which imaged the captured image as follows, and the pigment
- KOH potassium hydroxide
- centrifugation was performed at room temperature at 8000 rpm for 5 minutes, and the cells were precipitated at the bottom of the microcentrifuge tube to collect the cells.
- the supernatant in the microcentrifuge tube was removed, the cell pellet was immersed in liquid nitrogen together with the microcentrifuge tube, and the cell pellet was crushed with a mixer for 30 seconds.
- 1 ml of DMSO Dimethyl sulfoxide
- the mixture was centrifuged at 8000 rpm for 5 minutes at room temperature, and the supernatant (dye extract) was transferred to a new microcentrifuge tube to collect astaxanthin as a dye.
- the absorption of light at 492 nm and 750 nm of this astaxanthin dye extract was measured as absorbance by a spectrophotometer.
- 4.5 is a proportionality constant
- a 492 is the absorbance at 492 nm
- a 750 is the absorbance at 750 nm.
- the amount of astaxanthin obtained at this time was divided by the number of cells used for extraction of astaxanthin to calculate the amount of astaxanthin per cell.
- the pigment concentration of astaxanthin was determined from samples obtained by capturing a plurality of captured images having different brightness ratios in different captured images. Then, the average value of the luminance values of the R pixel and the B pixel in the cell image is obtained, and a regression equation indicating the correspondence between the luminance ratio that is the ratio of the average value and the extracted amount of astaxanthin, that is, the amount of astaxanthin is estimated.
- the regression equation was obtained by regression analysis for obtaining the correlation in FIG.
- FIG. 6 shows a luminance which is a ratio between the amount of chlorophyll pigment accumulated in the observation target cell (hematococcus) per cell and the average value of the luminance values of the G pixel and the B pixel in the pixel in the cell image. It is a graph which shows the regression type which shows a response
- the vertical axis indicates the amount of chlorophyll accumulated per cell (pg: picogram), and the horizontal axis indicates the average value of the luminance values of the G pixels of the pixels in the cell image divided by the average value of the B pixels.
- the luminance value ratio is shown.
- the regression equation for calculating chlorophyll is obtained in advance as follows, and is written and stored in the storage unit 22 in advance. Perform the above-described image analysis of the captured image to obtain numerical information in the captured image described above (other than the accumulated amount of pigment), and the average value of the luminance values of the G pixels and the average value of the luminance values of the B pixels in the cell image The luminance ratio is obtained.
- dye was extracted from the cell which imaged the captured image as follows, and the pigment
- DMSO 1 ml of DMSO was added to the microcentrifuge tube, and chlorophyll as a pigment was extracted by vortexing for 15 minutes.
- the mixture was centrifuged at 8000 rpm for 5 minutes at room temperature, and the supernatant (dye extract) was transferred to a new microcentrifuge tube to collect chlorophyll as a dye. Then, 1 ml of DMSO was added to the microcentrifuge tube until the cell pellet became white, and the extraction process of chlorophyll for centrifugal separation was repeated by vortexing for 15 minutes. Then, the absorption of light at 649 nm, 665 nm, and 750 nm of the chlorophyll pigment extract was measured as absorbance by a spectrophotometer. In the absorbance measurement results, the chlorophyll pigment concentration was determined by the following equation.
- Chlorophyll amount ( ⁇ g / ml) 14.85 (A 665 -A 750 ) -5.14 (A 649 -A 750 )
- 14.85 and 5.14 are proportional constants
- a 665 is the absorbance at 665 nm
- a 649 is the absorbance at 649 nm
- a 750 is the absorbance at 750 nm.
- the amount of chlorophyll per cell was calculated by dividing the amount of chlorophyll obtained at this time by the number of cells used for extraction of chlorophyll.
- the pigment concentration of chlorophyll was determined from samples obtained by capturing a plurality of captured images having different brightness ratios in different captured images. Then, an average value of the luminance values of the G pixel and the B pixel in the cell image is obtained, and a regression equation indicating the correspondence between the luminance ratio, which is the ratio of the average values, and the extracted amount of chlorophyll, that is, the amount of chlorophyll is estimated.
- the regression equation was obtained by regression analysis for obtaining the correlation in FIG.
- hematococcus cell culture was performed as follows. That is, a sample for quantification of the amount of astaxanthin for obtaining a micrograph of a captured image of Haematococcus and a regression equation was prepared as follows. In addition, hematococcus cell culture in the medium was performed for 3 months under continuous light. Then, 10 ml of the culture solution of Haematococcus pluvialis strain K0084 was inoculated against 90 ml of the medium inserted in a 300 ml flask.
- the culture was performed by standing under white light having a photon flux density of 45 ⁇ E (Einstein) m ⁇ 2 s ⁇ 1 at an ambient temperature of 25 ° C. Then, immediately after the start of culture, 30 ⁇ l was taken from the flask on the 7th and 14th days, and observed with a microscope and captured images.
- the medium for culturing Haematococcus in this embodiment is 4.055 mM KNO 3 , 0.347 mM CaCl 2 , 0.189 mM Na 2 CO 3 , 0.304 mM MgSO 4 , 0.175 mM K 2 HPO.
- the microscopic observation of Haematococcus in the culture medium was performed as follows. On the surface of a 1 mm thick slide glass, a square frame having a side of 1 cm was formed by nail enamel, and 30 ⁇ l of a hematococcus culture solution was dropped into the frame. Then, using a 0.17 mm thick cover glass, a nail enamel frame was sealed with nail enamel to prepare a preparation.
- Microscope observation uses an upright microscope equipped with a 40x objective lens, and a microscope with a color CCD (Charge Coupled Device) camera uses a resolution of 2040 ⁇ 1536 pixels (number of pixels: about 3 million pixels). It was.
- the captured image was stored as a JPEG (Joint Photographic Experts Group) image in RGB format (R pixel, G pixel, and B pixel) in the image storage unit 21 via the control unit 11.
- JPEG Joint Photographic Experts Group
- FIG. 7 is a flowchart showing an operation example of cell observation processing in the cell observation device 1 according to the present embodiment.
- FIG. 8 is a diagram showing the outermost edge portion of the cell image in which the edge image and the plastid image are superimposed.
- FIG. 8A shows a captured image that is a bright field image captured by the imaging apparatus 100. In this embodiment, no pretreatment such as fluorescent staining is performed on the cells.
- FIG. 8B shows a chromophoric image generated by the plastid region extraction unit 14.
- FIG. 8C shows the cell region image extracted by the contour extraction unit 17.
- FIG. 8D is an image obtained by superposing the plastid image and the cell region image, and shows the outermost edge portion of the cell image.
- the observer prepares a culture medium preparation, and captures an image of hematococcus cells in the preparation using a CCD camera.
- the observer searches for a region of the field of view where hematococcus cells do not overlap in the vertical direction and are not in contact with other cells. That is, a region where hematococcus cells are arranged in a single layer without contacting other cells on the preparation surface is searched.
- hematococcus is imaged while avoiding areas where it is clearly determined that there are abnormal cells or foreign bodies, and areas where the slide glass and wall-glass are dirty and the background around the cells is uneven. Further, a plurality of these fields of view are extracted, and a captured image is captured while moving the stage of the microscope in a certain direction so as not to capture the same cell a plurality of times (0.5 seconds / sheet). For example, the number of imaging layers from which information on the number of cells of 200 can be obtained is imaged.
- Step S1 The observer inputs a control signal from the input means (for example, a keyboard), not shown, to adjust the hue of the cell image in the captured image stored in the image storage unit 21 to the cell observation device 1.
- the control unit 11 reads the captured image data ((a) in FIG. 8) from the image storage unit 21 and sends the read captured image to the hue adjustment unit 12.
- the hue adjustment unit 12 adds up the histograms of the luminance values of the R pixel, the G pixel, and the B pixel of the captured image, and selects the luminance of the mode value as the luminance value of the background.
- the hue adjustment unit 12 sets the luminance values of the R pixel, the G pixel, and the B pixel of the captured image so that the background area other than the cell image that is the cell image is gray. The brightness value of each pixel is adjusted. Then, the control unit 11 supplies the captured image in which the R pixel, the G pixel, and the B pixel are adjusted to the image display unit 24 and the contour extraction unit 13 and writes the image in the image storage unit 21.
- the image processing from step S2 below is performed using the captured image after the background area is adjusted to gray.
- Step S2 The contour extraction unit 13 extracts R pixel data from RGB pixel data constituting one pixel of the captured image. Then, the contour extracting unit 13 adjusts the luminance value so that the minimum and maximum luminance values are 0 and 255 for the extracted R pixel data.
- the contour extracting unit 13 extracts pixels of the edge (contour) of the shape of the image in the captured image by the Canny method using the adjusted R pixel data.
- the R pixel is used because, for example, the cell membrane of Hematococcus, which is a microalgae, is red, which is convenient for extracting the edge of the cell image in the captured image.
- the contour extracting unit 13 outputs the extracted edge image to the image dividing unit 15.
- Step S3 When the edge extraction in the contour extraction unit 13 ends, the control unit 11 supplies the captured image supplied to the contour extraction unit 13 to the plastid region extraction unit 14.
- the plastid region extraction unit 14 extracts B pixel data from RGB pixel data constituting one pixel of the captured image. Then, the plastid region extraction unit 14 binarizes the luminance value of each B pixel in the captured image by the Otsu method, and is a plastid region that is a region of intracellular plastids (astaxanthin and chlorophyll plastids). Extract pixels of.
- the chromophoric region extracting unit 14 outputs the chromophoric region that has become black due to binarization to the image dividing unit 15 as a chromophoric image (FIG. 8B).
- Step S4 The image dividing unit 15 superimposes the edge image supplied from the contour extracting unit 13 and the chromophoric image supplied from the chromophoric region extracting unit 14, and deletes an excess edge portion from the edge image.
- a new edge image is generated by supplementing the cell shape which is insufficient.
- the image segmentation unit 15 uses the extracted edge image information to extract the captured image segmentation (region) using the extracted edge image information in order to extract the cell image region in the captured image that is the object to be extracted. Split). Further, the image dividing unit 15 thins the boundary line of the segmented region, that is, generates a segment boundary line (a boundary line having a width of one pixel) in the captured image.
- Step S5 The image composition unit 16 binarizes the luminance values of RGB pixels, that is, R pixels, G pixels, and B pixels constituting one pixel in the captured image by the Otsu method described above. Then, the image composition unit 16 extracts a pixel set region that is a region of pixels in which at least one luminance value is equal to or less than a threshold value among RGB pixels that are binarized to form one pixel. Actually, this pixel collection region is included in the region of the cell image in the captured image.
- Step S6 The image composition unit 16 superimposes the boundary lines of the segments on the image of the pixel set area in the captured image, and obtains the above-described ratio of the pixel set area for each segment. Then, the image composition unit 16 selects a segment in which the ratio of the pixel collection area in the segment exceeds a preset ratio as a cell segment that is an area of a cell image in the captured image. At this point, a background region that is not yet an actual cell region may be included.
- Step S7 the image composition unit 16 calculates the average value and variance of the pixel values of each of the R pixel, G pixel, and B pixel in the cell segment for each selected cell segment. Then, the image composition unit 16 selects a cell segment having a large variance in luminance value of at least one of the R pixel, the G pixel, and the B pixel as a cell corresponding to the captured region of the cell in the captured image. Select as a segment.
- the image composition unit 16 sets a segment other than the cell segment selected as the cell segment as the background segment.
- the image composition unit 16 synthesizes the cell segment and the background segment, forms a cell image and a background image, and forms a composite image.
- the above-described classification process between the cell segment and the background segment is performed by utilizing the fact that the luminance value dispersion in the cell segment is larger than that in the background segment.
- Step S8 Next, in order to select an image area to be a cell image in the synthesized image, the cell area extraction unit 17 divides the image area to be a cell image from the synthesized image as a segment area by the water-shed method already described. .
- Step S9 the cell region extraction unit 17 sorts the segment that is in contact with the edge of the captured image among the divided segments as a Touch. Further, the remaining segments are sorted into a circular segment close to the circular model image and a non-circular segment different from the circular model image using Chordiogram. A circular segment close to this circle is recognized as a microalgal cell, that is, hematococcus. In addition, the cell region extraction unit 17 integrates non-circular segments other than the circular segments, and outputs them to the cell structure extraction unit as a final cell region image ((d) in FIG. 8).
- Step S10 The cell structure extraction unit 18 calculates the numerical information shown in FIG. 3 in the cell image in the captured image already described, and writes and stores it in the table storage unit 23 for each cell image.
- the cell shape detection unit 19 extracts the luminance values of the R pixel and the G pixel in the region detected by the plastid region extraction unit 14 as the plastid region in the cell image. Then, when both the luminance values are less than a preset threshold value, the cell shape detection unit 19 determines that the cell is an amber mold, and “mixed other species organism” with respect to the Type in the numerical information of FIG. 3 of the cell. ".
- Step S11 The control unit 11 determines whether or not the image analysis processing has been completed for all captured images in the image storage unit 21. At this time, when the image analysis processing for all the captured images in the image storage unit 21 is completed, the control unit 11 advances the processing to step S12. On the other hand, if the image analysis processing for all the captured images in the image storage unit 21 has not been completed, the control unit 11 returns the process to step S1, reads out a new captured image from the image storage unit 21, and performs image analysis processing. Continue.
- Step S12 The cell structure extraction unit 18 reads out the numerical information from the numerical table shown in FIG. 3 stored in the table storage unit 23, calculates the numerical information of the numerical information Continant, Algae, Other, and Cell count shown in FIG. Write and store for each folder of target cells.
- this folder for example, numerical information obtained from cell images in a plurality of captured images used for calculation of pigment amounts of astaxanthin and chlorophyll at a predetermined date and time is stored.
- the edge image extracted from the captured image that is a microphotograph of a cell such as a microalgae and the plastid image are superimposed, and the image of the cell that is insufficient in the edge image by the plastid image.
- the outer shape of the cell image in the captured image can be extracted with higher accuracy than before, and the overall shape of the cell and the plastid region in the cell (for example, the region of the produced substance to be produced) ) Ratio can be easily detected.
- the control unit 11 causes the pigment value extraction unit 20 to perform pigment amount extraction processing.
- the pigment value extraction unit 20 stores OuterRedIntensity (average value of R pixel luminance values in the cell image) and OuterBlueIntensity (average value of B pixel luminance values in the cell image) from the table storage unit 23 in the folder. Read from the numerical information of cell images of all target algae. Next, the dye value extraction unit 20 adds the OuterRed Intensity of the whole cell image to obtain the OuterRed Intensity addition value, and similarly adds the OuterBlue Intensity of the whole cell image to obtain the OuterBlue Intensity addition value.
- the pigment value extraction unit 20 divides the OuterRed Intensity addition value by the OuterBlue Intensity addition value, and calculates an Astaxanthin predictor that is a luminance ratio of the R pixel and the B pixel.
- the pigment value extraction unit 20 writes and stores the calculated Astaxanthin predictor as the numerical information of FIG. 4 in the table storage unit 23.
- the pigment value extraction unit 20 reads the regression equation for calculating astaxanthin from the storage unit 22, substitutes the obtained Astaxanthin predictor for this regression equation, obtains the accumulated amount of astaxanthin per cell, and stores it in the table as Astaxanthin It is written and stored as numerical information in FIG.
- the pigment value extraction unit 20 stores OuterGreenIntensity (average value of luminance values of G pixels in the cell image) and OuterBlueIntensity (average value of luminance values of B pixels in the cell image) from the table storage unit 23. It reads out from the numerical information of the cell image of all the target algae inside.
- the dye value extraction unit 20 adds the OuterGreen Intensity of the whole cell image to obtain the OuterGreen Intensity addition value, and similarly adds the OuterBlue Intensity of the whole cell image to obtain the OuterBlue Intensity addition value.
- the pigment value extraction unit 20 divides the OuterGreen Intensity addition value by the OuterBlue Intensity addition value, and calculates a Chlorophyll predictor that is a luminance ratio of the R pixel and the B pixel.
- the pigment extraction unit 19 writes and stores the calculated Chlorophyll predicator as the numerical information of FIG. 4 in the table storage unit 23.
- the pigment value extraction unit 20 reads the regression equation for calculating chlorophyll from the storage unit 22, substitutes the obtained Chlorophyll predicator into this regression equation, obtains the accumulated amount of astaxanthin per cell, and stores it in the table as Chlorophyll. It is written and stored as numerical information in FIG.
- FIG. 9 is a graph showing a change in the accumulation amount of astaxanthin and chlorophyll measured by the cell observation device 1 after inoculating Haematococcus on a fresh medium.
- the vertical axis represents the amount of pigment (accumulated amount) per cell
- the horizontal axis represents the culture period (weeks).
- the amount of astaxanthin and chlorophyll accumulated increased immediately after inoculation (week 0), week 1, and week 2 after inoculation of hematococcus to the medium.
- This result is consistent with the finding that in actual hematococcus cells, the accumulated amount of astaxanthin reaches a plateau after one week, and the accumulated amount of chlorophyll increases with time. Therefore, the cell observation apparatus 1 according to the present embodiment can accurately estimate the accumulated amount of the pigment in the cultured state without actually destroying the cell and measuring the accumulated amount of the plastid. .
- FIG. 10 is a diagram showing correspondence between numerical information (OuterAxisRatio (L / S), OuterArea) of cell images obtained by the cell observation apparatus 1 from image analysis, and classification of cells visually.
- the vertical axis represents the ratio of the major axis and minor axis of the cell image OuterAxisRatio (L / S), and the horizontal axis represents the area OuterArea of the cell image.
- the “+” mark is a cell visually determined as a zoospore
- the “ ⁇ ” mark is a cell visually determined as a transition cell during the transition from the zoospore to the Palmera cell
- the “ ⁇ ” mark is It is a cell that has been visually determined as a Palmera cell (a hematococcus cell that produces astaxanthin as a pigment).
- FIG. 10 is a diagram for finding conditions for discriminating between zoospores and palmera cells having different physiological states.
- hematococcus cells After being transferred to the fresh medium, hematococcus cells were classified into three types of cells that were in a transitional state that could not be discriminated from zoospores and palmera cells in 20 captured images.
- the captured image was analyzed by the cell observation device 1 of the present embodiment, and quanta of 138 types of cells determined to be hematococcus cells were used to discriminate between zoospores and Palmera cells.
- quanta of 138 types of cells determined to be hematococcus cells were used to discriminate between zoospores and Palmera cells.
- zoospores are represented by “+”
- Palmera cells are represented by “ ⁇ ”
- transitional cells are represented by “O”.
- a cell having an area of 3000 pixels or more can be determined as a Palmera cell, and a pixel having a cell area of less than 3000 can be determined as a zoospore.
- a cell having a cell area of less than 3000 pixels and a long / short axis ratio (L / S) of 1.05 or more is a zoospore.
- cells having a cell area of 3000 pixels or more and a long / short axis ratio (L / S) of less than 1.05 are regarded as Palmera cells. Also from this result, it can be seen that the discrimination of the cell observation device 1 according to the present embodiment is the same as that of the observer's knowledge.
- hematococcus has been described as an example of a cell to be observed.
- the present invention is not limited to this hematococcus, and any cell can be applied as an observation target as long as it is a single cell.
- any cell can be applied as an observation target as long as it is a single cell.
- characteristic forms in other cells such as Chlorella, Nanochloropsis, Donariella, Botryococcus can be quantified.
- Astaxanthin is a kind of carotenoid.
- FIG. 5 has already described the estimation of the amount of astaxanthin by a regression equation using the average value of the luminance values of the R pixel and the B pixel in the cell image. Further, in FIG.
- the estimation of the chlorophyll amount is described by a regression equation using the luminance values of the G pixel and the B pixel in the cell image.
- the average value of the luminance value of the G pixel in the cell image is added to the average value of the luminance value of the R pixel and the B pixel in the cell image, and
- Each of the amount of astaxanthin and the amount of chlorophyll is determined using a multiple regression equation for determining the amount of pigment produced by the cell from the average value of the luminance values of each of the R pixel, G pixel, and B pixel.
- the multiple regression equation for obtaining the amount of astaxanthin and the multiple regression equation for obtaining the amount of chlorophyll are different multiple regression equations (described later), and are written and stored in the storage unit 22 in advance.
- the pigment value extraction unit 20 already described performs estimation of the pigment amount using this multiple regression equation.
- the pigment value extraction unit 20 reads the multiple regression equation for calculating astaxanthin from the storage unit 22 and calculates the average value of the luminance values of the R pixel, the G pixel, and the B pixel in the cell image as the weight of each of the astaxanthin and chlorophyll. Substituting into the regression equation, the accumulated amount of astaxanthin or chlorophyll per cell (astaxanthin amount, chlorophyll amount) is obtained, and is written and stored as numerical information of FIG. 4 in the table storage unit 23 as Astaxanthin or Chlorophyll.
- FIG. 11 shows the amount of pigment per ml of culture solution of chlorophyll accumulated in the actually observed cell (hematococcus) and the average value of the luminance values of the R pixel, G pixel, and B pixel in the pixel in the cell image. It is a graph which shows the correlation with the pigment amount estimated from the used multiple regression equation.
- the vertical axis shows the amount of chlorophyll per 1 ml of the culture solution ( ⁇ g / ml) converted from the accumulated amount of chlorophyll per cell (pg: picogram) estimated by the multiple regression equation, and the horizontal axis is actually measured.
- the amount of chlorophyll per ml of culture broth ( ⁇ g / ml) is shown.
- the correlation determination coefficient R 2 in FIG. 11 is 0.999.
- the multiple regression equation for estimating the amount of chlorophyll is a multiple regression analysis that correlates the correspondence between the luminance values of the R pixel, G pixel, and B pixel in the pixel in the cell image and the measured amount of chlorophyll per cell. Generated by.
- the R pixel and G pixel obtained from the measured value of the chlorophyll amount of the observation target cell (cultured cell cultured under LL conditions) used as training data and the cell image of the observation target cell used as this training data. And the correlation with the estimated amount of chlorophyll obtained from the luminance value of the B pixel using a multiple regression equation. Moreover, about the measurement of the chlorophyll amount per 1 ml culture solution, the process which extracts the chlorophyll similar to the process in description of FIG. 6 was performed, chlorophyll was extracted from the cell, and the chlorophyll amount per 1 ml culture solution was calculated
- the LL condition is a culture condition in which cells are always placed in a bright environment at a predetermined illuminance when cells are cultured.
- FIG. 12 shows the measured amount of pigment per ml of culture solution of chlorophyll accumulated in the observation target cells (hematococcus) and the average value of the luminance values of the R pixel, G pixel, and B pixel in the pixels in the cell image. It is a graph which shows the correlation with the pigment amount estimated from the used multiple regression equation.
- the horizontal axis represents the actually measured chlorophyll amount ( ⁇ g / ml) per 1 ml of the culture solution.
- the coefficient of determination R 2 of the correlation for the FIG. 12 is a 0.963.
- the multiple regression equation for estimating the amount of chlorophyll is a multiple regression analysis that correlates the correspondence between the luminance values of the R pixel, G pixel, and B pixel in the pixel in the cell image and the measured amount of chlorophyll per cell. Generated by.
- the R pixel and the G pixel obtained from the measured value of the chlorophyll amount of the observation target cell (cultured cell cultured under the LD condition) used as the test data and the cell image of the observation target cell used as the test data. And the correlation with the estimated amount of chlorophyll obtained from the luminance value of the B pixel using a multiple regression equation. Moreover, about the measurement of the chlorophyll amount per 1 ml culture solution, the process which extracts the chlorophyll similar to the process in description of FIG. 6 was performed, chlorophyll was extracted from the cell, and the chlorophyll amount per 1 ml culture solution was calculated
- the LD condition is a culture condition in which when a cell is cultured, a bright state with a predetermined illuminance and a dark state with a predetermined illuminance are placed in an environment that is repeated in a preset cycle.
- FIG. 12 shows, as described above, a multiple regression equation in which the cells cultured under the LL conditions described in FIG. 11 are generated as training data, and the cells cultured under the LD conditions are used as test data. The amount is being measured.
- Each of the LL condition and the LD condition is a completely different group as a statistical population of cultured cells with different culture conditions.
- FIG. 13 shows the average amount of the pigment amount per 1 ml of the culture liquid of astaxanthin accumulated in the actually observed cell (hematococcus) and the luminance value of the R pixel, G pixel, and B pixel in the pixel in the cell image. It is a graph which shows the correlation with the pigment amount of astaxanthin estimated from the used multiple regression equation.
- the graph of FIG. 13 shows the amount of astaxanthin ( ⁇ g / ml) per 1 ml of the culture broth converted from the accumulated amount of astaxanthin per cell (pg: picogram) estimated on the vertical axis, and the horizontal axis represents the measured per 1 ml of culture broth.
- the amount of astaxanthin ( ⁇ g / ml) is shown.
- the correlation determination coefficient R 2 in FIG. 13 is 0.963.
- R pixels and G pixels obtained from the actual measurement value of the amount of astaxanthin of the observation target cells (cultured cells cultured under LL conditions) used as training data, and the cell image of the observation target cells used as the training data. And the correlation with the estimated amount calculated
- the LL condition is a culture condition in which cells are always placed in a bright environment at a predetermined illuminance when cells are cultured.
- FIG. 14 shows the measured amounts of astaxanthin per ml of culture solution of astaxanthin accumulated in the actually observed cells (hematococcus) and the average values of the luminance values of the R, G, and B pixels in the pixels in the cell image. It is a graph which shows the correlation with the pigment amount of astaxanthin estimated from the used multiple regression equation. As in the graph of FIG. 13, the graph of FIG. 14 shows the amount of astaxanthin ( ⁇ g / ml) per 1 ml of the culture solution converted from the amount of astaxanthin accumulated per cell (pg: picogram) estimated on the vertical axis. The axis indicates the amount of astaxanthin ( ⁇ g / ml) per 1 ml of the measured culture solution.
- the correlation determination coefficient R 2 in FIG. 14 is 0.975.
- the LD condition is a culture condition in which when a cell is cultured, a bright state with a predetermined illuminance and a dark state with a predetermined illuminance are placed in an environment that is repeated in a preset cycle.
- FIG. 14 shows, as described above, a multiple regression equation in which cells cultured under the LL conditions described in FIG. 13 are generated as training data, and cells cultured under the LD conditions as test data. The amount is being measured.
- Each of the LL condition and the LD condition is a completely different group as a statistical population of cultured cells with different culture conditions.
- FIG. 15 is a diagram for explaining an equation for multiple regression analysis used to estimate each of the chlorophyll amount and the carotenoid amount.
- the vertical axis represents the dye amount ( ⁇ g / ml) converted per 1 ml of the culture solution, and the horizontal axis represents the time (days) elapsed since the start of the culture.
- black circles are actually measured chlorophyll amounts, and white circles are chlorophyll amounts estimated by a multiple regression equation.
- the black triangle is the measured value of the carotenoid amount actually measured, and the white triangle is the carotenoid amount estimated by the multiple regression equation.
- the multiple regression equation (also shown in FIG. 15) of the chlorophyll concentration (pg / cell) shown below is an equation for estimating the amount of chlorophyll contained per cell.
- Y chl ⁇ 0.46 ⁇ I R + 0.56 ⁇ I G ⁇ 0.83 ⁇ I B ⁇ 72.01
- the multiple regression equation (also shown in FIG. 15) of the carotenoid concentration (pg / cell) shown below is an equation for estimating the amount of carotenoid contained per cell.
- Y car 0.75 ⁇ I R ⁇ 0.22 ⁇ I G ⁇ 0.27 ⁇ I B ⁇ 61.83
- Y chl indicates the chlorophyll concentration
- Y car indicates the carotenoid concentration
- I R is the average value of the luminance values of the red channel in plastids in cells image.
- IG is the average value of the luminance values of the green channel in the plastids in the cell image.
- I B is the average value of the luminance values of the blue channel in the plastids in a cell image.
- the coefficient a is a coefficient by which the average value of the luminance values of the red channel of the pigment body pixels is multiplied.
- the coefficient b is a coefficient by which the average value of the luminance values of the green channel of the pigment body pixels is multiplied.
- the coefficient c is a coefficient by which the average value of the luminance values of the blue channel of the pigment body pixels is multiplied.
- the coefficient d is a constant in the multiple regression equation.
- ⁇ Distinction between Haematococcus zoospore and palmeroid by random forest method basically, a machine model based on random forest method is used to generate a tree model for clustering that separates each cell indicated by the cell image included in the captured image into a zoospore and a palmeroid. . Then, the generated tree model is written and stored in the storage unit 22 in advance.
- palmeroid is also called palmera cell.
- the cell morphology detection unit 19 reads out the tree model from the storage unit 22 and performs clustering of cell images included in the captured image based on the tree model. Further, the cell morphology detection unit 19 displays the clustering result on the display unit 24 as a clustering image (for example, the image of FIG. 20 described later), and visually shows the ratio of the zoospore to the palmeroid at that time. Let them express. By looking at this clustering image, the ratio between the zoospore and the palmeroid at that time can be easily estimated visually. Further, the cell shape detection unit 19 may be configured to repeatedly perform machine learning by the random forest method and update the tree model by inputting a correct answer of the clustering result.
- a clustering image for example, the image of FIG. 20 described later
- FIG. 16 is a diagram illustrating captured images obtained by capturing each of the zoospore and the palmeroid.
- cells indicated by black triangles are palmeroids
- cells indicated by arrows are zoospores. That is, each of the cells C_1, C_2, C_3, C_4, and C_7 is a zoospore, and the cells C_5 and C_6 are palmeroids.
- palmeroids and zoospores are easy because the zoospores have flagella etc., while humans can identify all of the large number of cells in culture and migrate with palmeroids. It is difficult to detect the ratio of the number of children.
- not only numerical clustering of each cultured cell by the tree model, but also the result of clustering is visually displayed on the display device to clearly notify the ratio of zoospore to palmeroid be able to.
- FIG. 17 is a diagram showing a result of clustering the cultured cells of the training data using a tree model generated using the two parameters of the cultured cells of the training data by the random forest method.
- the two parameters are the ratio of the major axis and minor axis OuterAxisRatio (L / S), which is numerical information of the cell image obtained from the image analysis by the cell observation device 1, and the area OuterArea of the cell image.
- the vertical axis represents the ratio of the major axis and the minor axis of the cell image OuterAxisRatio (L / S)
- the horizontal axis represents the area OuterArea of the cell image.
- the discrimination rate of the training data cells by the tree model generated from the training data cells is 100%.
- the discrimination rate indicates a correct answer rate of discrimination in which the discrimination result discriminates a zoospore as a zoospore and discriminates a palmeroid as a palmeroid.
- white circles indicate palmeroid cells
- white diamonds indicate zoospore cells.
- FIG. 18 is a diagram illustrating a result of clustering the cultured cells of the test data using a tree model generated by using the two parameters of the cultured cells of the training data by the random forest method.
- the two parameters are the ratio of the major axis and minor axis OuterAxisRatio (L / S), which is numerical information of the cell image obtained from the image analysis by the cell observation device 1, and the area OuterArea of the cell image.
- the vertical axis represents the ratio of the major axis and the minor axis of the cell image OuterAxisRatio (L / S)
- the horizontal axis represents the area OuterArea of the cell image.
- each cell of the training data and the test data is composed of hematococcus cells on the second or third day transferred to a fresh medium as a population, and 341 cells extracted from this population by 341 cells each.
- FIG. 19 is a diagram showing a result of clustering the cultured cells of the training data using a tree model generated using the 25 parameters of the cultured cells of the training data by random forest method machine learning. 25 parameters are added to each of the ratio of the major axis and minor axis OuterAxisRatio (L / S), which is numerical information of the cell image obtained from the image analysis by the cell observation device 1, and the area OuterArea of the cell image. And each of the parameters obtained from the image analysis by the cell observation device 1 described in FIG. In FIG.
- the vertical axis represents the ratio of the major axis and the minor axis of the cell image OuterAxisRatio (L / S), and the horizontal axis represents the area OuterArea of the cell image.
- the discrimination rate of the training data cells by the tree model generated from the training data cells is 100%.
- FIG. 20 is a diagram illustrating a result of clustering the cultured cells of the test data using a tree model generated using the 25 parameters of the cultured cells of the training data by random forest method machine learning.
- the 25 parameters are the ratio of the major axis and minor axis OuterAxisRatio (L / S), which is numerical information of the cell image obtained from the image analysis by the cell observation device 1, and the area OuterArea of the cell image.
- FIG. 20 as in FIG.
- the vertical axis represents the ratio of the major axis to the minor axis of the cell image, OuterAxisRatio (L / S), and the horizontal axis represents the area OuterArea of the cell image.
- the discrimination rate of the test data cells by the tree model generated from the training data cells is 89%.
- the discrimination rate between palmeroids and zoospores for the test data cells was 79%, but in the case of 25 parameters, the accuracy is improved by 89% and 10%.
- the determination is made after the culture on the second day or the third day after the start of the cell culture. Therefore, when the cell culture is performed for the first time using the above-described tree model in time series, a change in the ratio between the zoospore and the palmeroid is visually determined from the clustering result as shown in FIGS. Can be confirmed.
- the accuracy of clustering of zoospores and palmeroids can be further improved. It becomes easier to do with images.
- each of 25 parameters used for the machine learning in the random forest method in this embodiment is demonstrated.
- white circles indicate parameters extracted from cell images of cells cultured under LL conditions
- black circles are extracted from cell images of cells cultured under LD conditions.
- Each of the parameters shown is shown.
- the cell structure extraction unit 18 when extracting a parameter, the cell structure extraction unit 18 counts the number of pixels in a line segment to be measured in the cell image when the parameter is a unit of length, and the actual length is calculated from the counted number. The size is determined, that is, the number of pixels is converted into a numerical value having a length of ⁇ m.
- each numerical value from FIG. 21 to FIG. 45 is an average value obtained by averaging the measurement values of a plurality of cell images.
- FIG. 21 is a diagram showing a correspondence relationship between the incubation time and OuterOutlineLength (unit: ⁇ m) indicating the perimeter of the cell image of the cultured cell.
- the vertical axis indicates the measured value of OuterOutlineLength
- the horizontal axis indicates the number of days the cells are cultured (Day, the same applies to each of FIGS. 22 to 45 below).
- FIG. 22 is a diagram showing the correspondence between the culture time and OuterMaxRadius (unit: ⁇ m) indicating the maximum value of the distance from the coordinates of the center of gravity of the cell image of the cultured cell to the outer periphery of the cell image.
- the vertical axis represents the measured value of OuterMaxRadius
- the horizontal axis represents the number of days in which the cells were cultured.
- FIG. 23 is a diagram illustrating a correspondence relationship between the culture time and OuterArea (unit: ⁇ m 2 ) indicating the area of the cell image of the cultured cell.
- the vertical axis indicates the measured value of OuterArea
- the horizontal axis indicates the number of days in which the cells are cultured.
- FIG. 24 is a diagram showing a correspondence relationship between the culture time and the dimension Outer Short Axis Length (unit: ⁇ m) of the minimum width (short axis) portion of the cell image of the cultured cell.
- the vertical axis represents the measured value of Outer Short Axis Length
- the horizontal axis represents the number of days the cells were cultured.
- FIG. 25 is a diagram showing a correspondence relationship between the incubation time and OuterTotalRedIntensity (unit: color unit), which is a numerical value indicating the added value of the luminance value of the red channel (R pixel) of the whole cell in the cell image of the cultured cell. is there.
- the vertical axis represents the measured value of OuterTotalRedIntensity
- the horizontal axis represents the number of days the cells were cultured.
- the cell structure extraction unit 18 already described adds the luminance values of the R pixels in all the pixels in the cell image, and sets the addition result as OuterTotalRedIntensity.
- FIG. 26 is a diagram showing the correspondence between the culture time and InnerArea (unit: ⁇ m 2 ), which is the area in the plastid region of the cell image of the cultured cells.
- the vertical axis indicates the measured value of InnerArea
- the horizontal axis indicates the number of days in which the cells are cultured.
- FIG. 27 is a diagram showing a correspondence relationship between the incubation time and InnerRedRedIntensity (unit: color unit) which is a numerical value indicating the addition value of the luminance value of the red channel (R pixel) in the chromophoric region of the cell image.
- the vertical axis indicates the measured value of InnerTotalRedIntensity
- the horizontal axis indicates the number of days the cells are cultured.
- the already-explained cell structure extraction unit 18 adds the luminance values of the R pixels in all the pixels in the pigment body in the cell image, and sets the addition result as InnerTotalRedIntensity.
- FIG. 28 is a diagram showing a correspondence relationship between the incubation time and OuterTotalGreenIntensity (unit: color unit) which is a numerical value indicating the added value of the luminance value of the entire green channel (G pixel) in the cell image of the cultured cell. is there.
- the vertical axis indicates the measured value of OuterTotalGreenIntensity
- the horizontal axis indicates the number of days the cells are cultured.
- the already described cell structure extraction unit 18 adds the luminance values of the G pixels in all the pixels in the cell image, and the addition result is OuterTotalGreenIntensity.
- FIG. 29 is a diagram showing a correspondence relationship between the incubation time and Inner Mean Blue Intensity (unit: color unit), which is a numerical value indicating the average luminance value of the blue channel (B pixel) in the plastid region of the cell image.
- the vertical axis indicates the measured value of Inner Mean Blue Intensity
- the horizontal axis indicates the number of days the cells are cultured.
- the already described cell structure extraction unit 18 calculates the average value of the luminance values of the B pixels in all the pixels in the chromosomal body in the cell image, and sets the calculation result as Inner Mean Blue Intensity.
- FIG. 30 is a diagram showing a correspondence relationship between the culture time and DistanceFromCellCenterToInnerCenterOfMass (unit: ⁇ m), which is a numerical value indicating the distance from the center point of the cell image of the cultured cell to the center of gravity.
- the vertical axis represents the measured value of DistanceFromCellCenterToInnerCenterOfMass
- the horizontal axis represents the number of days on which the cells were cultured.
- the already-explained cell structure extraction unit 18 obtains the center coordinate value of the cell image and the centroid coordinate value of the centroid of the cell image from the pixel value of the cell image, obtains the difference between the center coordinate value and the centroid coordinate value, Let the difference be DistanceFromCellCenterToInnerCenterOfMass.
- FIG. 31 is a diagram showing a correspondence relationship between the incubation time and Outer Mean Blue Intensity (unit: color unit) that is a numerical value indicating the average luminance value of the blue channel (B pixel) in the cell image of the cultured cell.
- the vertical axis represents the measured value of Outer Mean Blue Intensity
- the horizontal axis represents the number of days for which the cells were cultured.
- the already described cell structure extraction unit 18 calculates the average value of the luminance values of the B pixels in all the pixels in the cell image, and sets the calculation result as OuterMeanBlueIntensity.
- FIG. 32 is a diagram showing a correspondence relationship between the incubation time and InnerTotalGreenIntensity (unit: color unit) that is a numerical value indicating the added value of the luminance value of the green channel (G pixel) in the plastid region of the cell image.
- the vertical axis represents the measured value of InnerTotalGreenIntensity
- the horizontal axis represents the number of days for which the cells were cultured.
- the already-described cell structure extraction unit 18 adds the luminance values of the G pixels in all the pixels in the pigment body in the cell image, and sets the addition result as InnerTotalGreenIntensity.
- FIG. 33 is a diagram showing a correspondence relationship between the culture time and OuterGreenGreenIntensity (unit: color unit) that is a numerical value indicating the average luminance value of the green channel (G pixel) in the cell image of the cultured cell.
- the vertical axis represents the measured value of Outer Mean Green Intensity
- the horizontal axis represents the number of days the cells were cultured.
- the already described cell structure extraction unit 18 calculates the average value of the luminance values of the G pixels in all the pixels in the cell image, and sets the calculation result as OuterGreenGreenIntensity.
- FIG. 34 is a diagram showing a correspondence relationship between the incubation time and AngleFromInnerCenterOfMassToFarEndOfLongAxis (unit: degree of angle) indicating the angle formed from the center of gravity of the cell image to the farthest coordinate point of the long axis.
- the vertical axis represents the measured value of AngleFromInnerCenterOfMassToFarEndOfLongAxis
- the horizontal axis represents the number of days the cells were cultured.
- the cell structure extraction unit 18 described above obtains the barycentric coordinate value of the center of gravity of the cell image from the pixel value of the cell image, extracts the coordinate value farthest from the received coordinate of the long axis of the cell image, The angle formed by the straight line connecting the coordinate values and the long axis of the cell image is detected, and this angle is defined as AngleFromInnerCenterOfMassToFarEndOfLongAxis.
- FIG. 35 is a diagram illustrating a correspondence relationship between the incubation time and OuterTotalBlueIntensity (unit: color unit) which is a numerical value indicating the addition value of the luminance value of the blue channel (B pixel) in the whole cell of the cell image of the cultured cell. is there.
- the vertical axis indicates the measured value of OuterTotalBlue Intensity
- the horizontal axis indicates the number of days the cells are cultured.
- the cell structure extraction unit 18 already described adds the luminance values of the B pixels in all the pixels in the cell image, and sets the addition result as OuterTotalBlueIntensity.
- FIG. 36 is a diagram showing the correspondence between the incubation time and InnerOutlineLength (unit: ⁇ m) indicating the perimeter of the plastids in the cell image.
- the vertical axis indicates the measured value of InnerOutlineLength
- the horizontal axis indicates the number of days the cells are cultured.
- FIG. 37 is a diagram showing a correspondence relationship between culture time and AreaRatio (unit: ⁇ m 2 / ⁇ m 2 ) indicating the ratio between the area of plastids in the cell and the area of the cell in the cell image.
- the vertical axis represents the measured value of AreaRatio
- the horizontal axis represents the number of days the cells were cultured.
- the already described cell structure extraction unit 18 calculates the ratio of the area of the plastid to the area of the entire cell image by dividing the area of the plastid by the area of the entire cell image from the pixel value of the cell image. Let the ratio be AreaRatio.
- FIG. 38 shows the correspondence between the incubation time and InnerTotalBlueIntensity (unit: color unit), which is a numerical value indicating the added value of the luminance value of the blue channel (B pixel) of the plastid in the cell image of the cultured cell.
- FIG. 38 the vertical axis represents the measured value of InnerTotalBlue Intensity, and the horizontal axis represents the number of days the cells were cultured.
- the already-described cell structure extraction unit 18 adds the luminance values of the B pixels in all pixels of the plastid region in the cell image, and sets the addition result as InnerTotalBlueIntensity.
- FIG. 39 is a diagram showing a correspondence relationship between incubation time and Outer Chordogram Distance (unit: relative frequency), which is a numerical value indicating the Chordogram distance (the parameter already described) in the cell image.
- the vertical axis represents the measurement value of Outer Chordiogram Distance
- the horizontal axis represents the number of days in which the cells were cultured.
- FIG. 40 is a diagram showing a correspondence relationship between the incubation time and OuterAxisRatio (L / S) (unit: ⁇ m / ⁇ m), which is a numerical value indicating the ratio between the long axis and the short axis of the cell image of the cultured cell.
- the vertical axis indicates the measured value of OuterAxisRatio (L / S)
- the horizontal axis indicates the number of days in which the cells are cultured.
- FIG. 41 is a diagram showing a correspondence relationship between the incubation time and Outer Mean Red Intensity (unit: color unit), which is a numerical value indicating the average luminance value of the red channel (R pixel) in the cell image of the cultured cell.
- the vertical axis represents the measured value of Outer Mean Red Intensity
- the horizontal axis represents the number of days in which the cells were cultured.
- the already described cell structure extraction unit 18 calculates the average value of the luminance values of the R pixels in all the pixels in the cell image, and sets the calculation result as OuterMeanRedIntensity.
- FIG. 42 is a diagram illustrating a correspondence relationship between the incubation time and OuterRoundfitness (unit: ⁇ m 2 / ⁇ m 2 ), which is a numerical value indicating the degree of conformity to the outer circle of the cell image.
- the vertical axis represents the measured value of OuterRoundfitness
- the horizontal axis represents the number of days the cells were cultured.
- FIG. 43 is a diagram showing a correspondence relationship between the incubation time and InnerMeanRedIntensity (unit: color unit), which is a numerical value indicating the average luminance value of the red channel (R pixel) of the plastid in the cell image of the cultured cell. is there.
- the vertical axis indicates the measured value of InnerMeanRedIntensity
- the horizontal axis indicates the number of days the cells are cultured.
- the already-explained cell structure extraction unit 18 calculates the average value of the luminance values of the R pixels in all the pixels in the plastids in the cell image, and sets the calculation result as InnerMeanRedIntensity.
- FIG. 44 is a diagram showing a correspondence relationship between the incubation time and InnerGreenGreenIntensity (unit: color unit), which is a numerical value indicating the average luminance value of the green channel (G pixel) of the plastid in the cell image of the cultured cell. is there.
- the vertical axis indicates the measured value of Inner Mean Green Intensity
- the horizontal axis indicates the number of days the cells are cultured.
- the cell structure extraction unit 18 already described calculates the average value of the luminance values of the G pixels in all the pixels in the plastid in the cell image, and the calculation result is Inner Mean Green Intensity.
- FIG. 45 is a diagram illustrating a correspondence relationship between the culture time and the dimension OuterLongAxisLength (unit: ⁇ m) of the maximum width (long axis) portion of the cell image of the cultured cell.
- the vertical axis represents the measured value of Outer Long Axis Length
- the horizontal axis represents the number of days the cells were cultured.
- the 25 parameters shown in FIGS. 21 to 45 are used to generate the tree model in the random forest method described above.
- the parameters extracted by the cell shape detection unit 19 including 25 from FIG. 21 to FIG. 45 evaluation for detecting temporal change and condition dependence of the cell shape in the growth process from hematococcus zoospore to palmeroid
- the parameters were extracted by principal component analysis.
- the contribution ratio of the first principal component PC1 was 51%.
- the first principal component PC1 includes eight parameters (including the long axis length OuterLongAxisLength and the cell area OuterArea) among the parameters extracted by the cell structure extraction unit 18 including the 25 elements shown in FIGS. It was found that there is a high correlation with (described later).
- the perimeter length OuterOutlineLength shown in FIG. 21 and the maximum width OuterMaxRadius from the center of gravity to the outer periphery of the cell image shown in FIG. 22 change linearly with time. And has a high correlation with the first principal component PC1.
- FIG. 46 is a diagram showing the correspondence between the first main component PC1 and the elapsed time since the start of culture.
- the vertical axis represents the score (score of PC1) of the first main component PC1
- the horizontal axis represents the elapsed time (Day) after the start of culture.
- the value of the first principal component decreases with time, but the state of change differs between the LL condition and the LD condition.
- the cells cultured under each of the LL condition and the LD condition have a correlation with the first principal component PC1, and naturally have a correlation with the time change.
- the parameter correlated with the first principal component PC1 can be used as a parameter for detecting the morphological change in the growth process of the observation target cell over time for both the LL condition and the LD condition. .
- FIG. 47 is a diagram for explaining eight parameters correlated with the first principal component PC1 (that is, correlated with the passage of time in the form change).
- the eight parameters are the long axis length OuterLongAxisLength (FIG. 45), the perimeter length of the cell OuterOutlineLength (FIG. 21), the maximum width from the center of gravity to the perimeter OuterMaxRadius (FIG. 22), and the cell area OuterArea (FIG. 23).
- the length of the short axis OuterShortAxisLength (FIG. 24), the added value of the luminance value of the red channel of the whole cell image OuterTotalRedIntensity (FIG. 25), the area of the intracellular dye part (plastid) InnerArea (FIG. 26), Each of the addition values InnerTotalRedIntensity (FIG. 27) of the luminance value of the red channel.
- FIG. 48 is a diagram showing the correspondence between generation and disappearance of flagella and the elapsed time since the start of culture.
- the vertical axis indicates the appearance rate (Fraction of flagellated cells), which is the frequency of appearance of cells with flagella
- the horizontal axis indicates the elapsed time (Day) after the start of culture. From FIG. 48, it can be seen that the flagellum generation peak in the cells is on the first day in the case of the cells in the culture environment under the LL condition and on the second day in the case of the cells in the culture environment under the LD condition. I understand.
- the appearance rate is obtained by dividing the number of cells having flagella by the total number of cells in the measurement region of the medium in which the cells are cultured.
- the average value of the brightness of the green channel in the cell shown in FIG. 33 differs from the 10th day after the start of cell culture. That is, in the culture under the environment of the LL condition, the value of Outer Mean Green Intensity in the cell image obtained by imaging the cell does not change. On the other hand, in the culture under the environment of the LD condition, the value of Outer Mean Green Intensity in the cell image obtained by imaging the cell is greatly reduced. This indicates that Outer Mean Green Intensity is a parameter that can separate cells cultured under the LD condition and cells cultured under the LL condition.
- the parameters having little correlation with the first principal component PC1 are temporally different in cell morphology during the growth depending on factors different from temporal changes in cell growth, that is, culture conditions and environments. Therefore, by extracting the parameters depending on the culture conditions and environment from the parameters having little correlation with the first principal component PC1, as described above, clustering of cells grown under each of the LL condition and the LD condition is performed. Is possible.
- parameters having the characteristics of the time-dependent change of the observation target cell and the conditions and environment-dependent changes can be extracted by principal component analysis.
- the dynamism of the temporal change in the growth of the observation target cell can be evaluated by the parameters extracted by the principal component analysis.
- discriminant analysis of different strains of chlorella cells was performed by discriminant analysis.
- a discriminant analysis was performed between a chlorella wild (WT) strain, a mutant (mutant) strain PkE6 derived from the wild strain, and a mutant strain PkE8.
- Discriminant analysis is used to determine which group (clustering) when new data is obtained when it is clear that each given data is separated into different groups. This is a method for obtaining a reference (discriminant function).
- the following processing is performed by the cell shape detection unit 19 using the parameter data shown in FIGS.
- FIG. 49 shows the appearance frequency, which is the frequency of appearance of each of the wild (WT) strain, mutant strain PkE6, and mutant strain PkE8 detected by an automatic cell counting device capable of simultaneously measuring the particle diameter and the number of particles.
- FIG. 49 the vertical axis indicates the appearance frequency, and the horizontal axis indicates the particle diameter of the cells. This particle diameter corresponds to the parameter OuterLongAxisLength in this embodiment.
- the particle size was confirmed to be large by the measurement using an automatic cell number measuring apparatus that the mutant strain PkE6 was large.
- WT is an abbreviation for wildtype.
- FIG. 50 shows the frequency of appearance for each particle size of each of the wild (WT) strain, the mutant strain PkE6, and the mutant strain PkE8, which is obtained by the cell morphology detection unit 19 in the cell observation device 1 according to the present embodiment.
- the vertical axis indicates the frequency of appearance
- the horizontal axis indicates the parameter OuterLongAxisLength (particle diameter) extracted from the cell image.
- the wild (WT) strain is indicated by a solid line
- the mutant strain PkE8 is indicated by a broken line
- the mutant strain PkE6 is indicated by a one-dot chain line.
- the appearance frequency is obtained by dividing the total probability density having each particle diameter by the total cells in the measurement region of the medium in which the cells are cultured.
- the probability density that each cell has a particle size was estimated by a kernel density estimation method that assumed that the obtained particle size of each cell was distributed according to a normal distribution.
- the particle size of the mutant strain PkE6 was confirmed by image analysis for each of the cell images in the captured image by the cell observation device 1 according to the present embodiment.
- the cell shape detection unit 19 detects the particle diameter of the cell image in a predetermined region in the captured image, and obtains the appearance frequency of the cell having this particle diameter for each particle diameter.
- FIG. 50 is a graph of this. As a result, it is understood that the mutant strain PkE6 is the largest, the mutant strain PkE8 is the next largest, and the wild strain (WT) is the smallest. Further, it can be seen from FIG. 50 that the mutant PkE8 also has a population of cells having the same size as that of the mutant PkE6.
- FIG. 51 is a table showing the results of the significant test of the parameters shown in FIGS.
- ID indicates the parameter name
- kw indicates the test result of the Kruskal-Wallis test
- e6 / wt indicates the result of the pair comparison of the U test between the mutant strain PkE6 and the wild strain (WT)
- e8 / wt shows the result of the U-comparison pair comparison between the mutant strain PkE8 and the wild strain (WT)
- e6 / e8 shows the result of the U-test pair comparison between the mutant strain PkE6 and the mutant strain PkE8.
- FIG. 52 is a diagram showing the discrimination results of wild (WT) strain, mutant strain PkE8, and mutant strain PkE6 based on the discriminant function LD1 and discriminant function LD2 obtained as a result of discriminant analysis.
- the vertical axis indicates the score of the discriminant function LD2
- the horizontal axis indicates the score of the discriminant function LD1.
- the white circle is the mutant strain PkE6, the black circle is the mutant strain PkE8, and the white square is the wild (WT) strain.
- the score of the discriminant function LD1 increases toward the + side, the color of the pigment in the cell becomes lighter.
- the discriminant function LD1 is highly correlated with a parameter indicating the luminance of each color channel in a pixel in a cell image or in a pigment body.
- the discriminant function LD2 is highly correlated with the parameter indicating the cell size of the cell image.
- FIG. 53 is a table showing the correlation coefficients between the discriminant functions LD1 and LD2 and the parameters shown in FIGS.
- ID indicates the parameter name
- the term LD1 indicates the correlation coefficient with the discriminant function LD1
- the term LD2 indicates the correlation coefficient with the discriminant function LD2.
- the discriminant function LD1 has a high correlation with the parameter indicating the luminance of the channel of each color in the pixel in the cell of the cell image or in the pigment body.
- the discriminant function LD2 is highly correlated with the parameter indicating the cell size of the cell image.
- FIG. 54 is a diagram showing the correspondence between the average luminance value (Outer Mean Intensity) of the cell pixel of each color channel having a high correlation with the discriminant function LD1, and the appearance rate of the cell.
- FIG. 54A shows the correspondence between the average value of the luminance value of the red channel of the pixel of the cell and the appearance rate of the cell having the average value.
- the vertical axis indicates the appearance rate of cells
- the horizontal axis indicates the numerical value of the parameter OuterMeanRedIntensity.
- the average brightness value of mutant PkE8 is the highest compared to wild (WT) and mutant PkE6.
- the average value of the luminance value of the mutant strain PkE6 has two peaks, that is, a high population as in the mutant strain PkE8 and a population lower than the wild (WT) strain and the mutant strain PkE8.
- the average value of the luminance value of the wild (WT) strain is lower than that of the mutant strain PkE8 and higher than that of the population having a lower luminance value of the mutant strain PkE6.
- FIG. 54 (b) shows the correspondence between the average value of the luminance value of the green channel of the pixel of the cell and the appearance rate of the cell having the average value.
- the vertical axis indicates the appearance rate of cells
- the horizontal axis indicates the numerical value of the parameter OuterGreenGreenIntensity.
- OuterMeanGreenIntensity the average value of the luminance value of the mutant strain PkE8 is the highest as compared to the wild (WT) strain and the mutant strain PkE6, as in OuterMeanRedIntensity.
- the average value of the luminance value of the mutant strain PkE6 has two peaks, that is, a high population as in the mutant strain PkE8 and a population lower than the wild (WT) strain and the mutant strain PkE8.
- the average value of the luminance value of the wild (WT) strain is lower than that of the mutant strain PkE8 and higher than that of the population having a lower luminance value of the mutant strain PkE6.
- (C) of FIG. 54 shows the correspondence between the average value of the luminance value of the blue channel of the pixel of the cell and the appearance rate of the cell having the average value.
- the vertical axis represents the appearance rate of cells
- the horizontal axis represents the numerical value of the parameter Outer Mean Blue Intensity.
- OuterMeanBlueIntensity as with OuterMeanRedIntensity, the average value of the luminance value of the mutant strain PkE8 is the highest compared to the wild (WT) strain and the mutant strain PkE6.
- the average brightness value of the mutant strain PkE6 is the lowest compared to the mutant strain PkE8 and the wild (WT) strain.
- the average brightness value of the wild (WT) strain is lower than that of the mutant strain PkE8 and higher than that of the mutant strain PkE6.
- the cell shape detection unit 19 estimates the probability density that each cell appears for each numerical value of the parameter OuterMeanRedIntensity for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6, and estimates The sum of the probability densities is divided by the number of cells in the entire strain to calculate the appearance rate. Similarly, the cell shape detection unit 19 estimates the probability density that each cell appears for each numerical value of the parameter OuterGreenGreenIntensity for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6, and the estimated probability The appearance rate is calculated by dividing the total density by the number of cells in the entire strain.
- the cell shape detection unit 19 estimates the probability density at which each cell appears for each numerical value of the parameter OuterMeanBlueIntensity for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6, and the estimated probability density Is divided by the number of cells in the entire strain to calculate the appearance rate.
- the estimation of the probability density that each cell has a particle size was performed by the kernel density estimation method.
- FIG. 55 is a diagram showing a correspondence between a parameter related to the size of a cell having a high correlation with the discriminant function LD2 and the appearance rate of the cell.
- FIG. 55A shows the correspondence between the length of the long axis OuterLongAxisLength and the appearance rate of cells having the length of the long axis OuterLongAxisLength.
- the vertical axis represents the appearance rate of cells
- the horizontal axis represents the value of the length of the long axis OuterLongAxisLength.
- the peak of the mutant strain PkE6 is the highest compared to the wild (WT) strain and the mutant strain PkE8.
- the peak of the wild (WT) strain is the smallest compared to the mutant strain PkE6 and the mutant strain PkE8.
- the peak of the mutant strain PkE8 is between the wild (WT) strain and the mutant strain PkE8.
- FIG. 55 (b) shows a correspondence between the short axis length Outer Short Axis Length and the appearance rate of cells having the short axis length Outer Short Axis Length.
- the vertical axis represents the appearance rate of cells
- the horizontal axis represents the numerical value of the length of the short axis OuterShortAxisLength.
- the mutant PkE6 has the highest peak compared to the wild (WT) and mutant PkE8.
- the peak of the mutant strain PkE8 is the smallest compared to the wild (WT) strain and the mutant strain PkE8.
- the peak of the wild (WT) strain is between mutant PkE6 and mutant PkE8.
- FIG. 55 (c) shows the correspondence between the cell area OuterArea and the appearance rate of cells having the cell area OuterArea.
- the vertical axis represents the cell appearance rate
- the horizontal axis represents the numerical value of the cell area OuterArea.
- the peak of the mutant strain PkE6 is the highest compared to the wild (WT) strain and the mutant strain PkE8.
- the peak of the mutant strain PkE8 is the smallest compared to the wild (WT) strain and the mutant strain PkE8.
- the peak of the wild (WT) strain is between mutant PkE6 and mutant PkE8.
- the cell shape detection unit 19 estimates the probability density at which each cell appears for each numerical value of the parameter OuterLongAxisLength for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6. The sum of the probability densities is divided by the number of cells in the entire strain to calculate the appearance rate. Similarly, the cell morphology detection unit 19 estimates the probability density that each cell appears for each numerical value of the parameter OuterShortAxisLength for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6, and the estimated probability The appearance rate is calculated by dividing the total density by the number of cells in the entire strain.
- the cell shape detection unit 19 estimates the probability density that each cell appears for each numerical value of the parameter OuterArea for each of the wild (WT) strain, the mutant strain PkE8, and the mutant strain PkE6, and the estimated probability density Is divided by the number of cells in the entire strain to calculate the appearance rate.
- the estimation of the probability density that each cell has a particle size was performed by the kernel density estimation method.
- the discriminant function LD1 correlated with the parameter related to the luminance value in the cell and the discriminant function LD2 correlated with the parameter related to the cell size
- the wild (WT ), Mutant PkE6 and mutant PkE8 cells can be separated into groups of respective strains.
- the program for realizing the function of the cell observation apparatus 1 in FIG. 1 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into the computer system and executed, thereby executing the cell.
- the physiological state may be detected and the culture may be managed.
- the “computer system” includes an OS and hardware such as peripheral devices.
- the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
- the “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM or a CD-ROM, and a hard disk incorporated in a computer system.
- the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line.
- a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included.
- the program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.
- a cell observation device that monitors each of the quantities in real time can be provided.
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Organic Chemistry (AREA)
- Zoology (AREA)
- Wood Science & Technology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Biotechnology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Immunology (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Microbiology (AREA)
- Proteomics, Peptides & Aminoacids (AREA)
- General Engineering & Computer Science (AREA)
- Genetics & Genomics (AREA)
- Pathology (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Medicinal Chemistry (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Sustainable Development (AREA)
- Biophysics (AREA)
- Geometry (AREA)
- Signal Processing (AREA)
- Dispersion Chemistry (AREA)
- Mathematical Physics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Apparatus Associated With Microorganisms And Enzymes (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Image Processing (AREA)
Abstract
Description
本願は、2012年11月28日に、日本に出願された特願2012-259880号に基づき優先権を主張し、その内容をここに援用する。
また、微細藻類は、種類によって、有用成分として炭化水素や必須不飽和脂肪酸(例えばDHA(Docosa Hexaenoic Acid)、EPA(Eicosa Pentaenoic Acid)など)、デンプン、色素などを生合成するものがあり、これらの生合成の機能の工業的利用が期待されている。
また、細胞を培養している培地中への他種生物の混入が微細藻類の細胞の生理状態に影響を及ぼす場合がある。この他種生物の混入による生理状態への影響が、細胞による有用成分の生産過程において問題となることも少なくない。
したがって、上述した培養中の培地へ混入した他種生物の検出も、微細藻類の細胞を用いた有用成分の高効率な生産のためには重要となる。
しかしながら、この非特許文献3による方法では、細胞の生理状態を多面的に判定できない。
また、特許文献2には、細胞形態定量値を用いる酵母の生理状態の評価方法が示されている。詳しくは、目的とする酵母の細胞の外郭、核、およびアクチン細胞骨格の蛍光染色画像を画像解析して、酵母細胞の形態学的な特徴に基づいて予め設定された形態パラメータについて、細胞形態定量解析値を求め、その値を予め用意したデータベースと比較することによって、目的とする酵母の生理状態を評価する方法である。
また、有用成分の蓄積量を把握する方法として、特許文献1では、Haematococcus pluvialisのアスタキサンチン量を、ジメチルスルフォキシドによって細胞から色素体を抽出し、492nmおよび750nmの吸光度を測定することで定量する方法が示されている。
しかし、測定には多数の微細藻類の細胞からの色素体の抽出を必要とし、色素体の抽出に時間を要する。
また、別の先行研究では、Haematococcus pluvialisに寄生するParaphysoderma sedebokerensisの遊走子嚢をFITC-WGAで染色している(非特許文献5)。
以上に述べたとおり、微細藻類の細胞の生育状況、また微細藻類の細胞の培地に混入する他種生物(例えば寄生菌)などの混入状況、微細藻類の細胞による有用物質の生産量の各々をリアルタイムにモニターする簡便な方法はない。
これにより、本発明によれば、細胞の生理状態の指標、他種生物の混入状況、有色色素体の蓄積状況が定量値として得られ、細胞の生理状態の把握や、有用成分の生産状況の把握が容易になる。
また、詳細は後に説明するが、ピクセル値で示された細胞の面積と、細胞画像の最外縁における最大幅(OuterLongAxisLength)を最小幅(OuterShortAxisLength)で除算した長軸短軸比(OuterLongAxisLengthとOuterShortAxisLengthとの比L/S)とがそれぞれの細胞画像に記述されている。この図2においては、細胞画像の中央には各細胞画像を識別する識別番号(ID)が記載されている。
また、細胞の発育の形態として、ヘマトコッカスの発育状態の初期である遊走子と、パルメラ細胞との顕微鏡により観察者が判定した結果が示されている。
色合い調節部12は画像記憶部から撮像画像を読み出し、当該撮像画像の各画素のRGBの輝度値(階調度)を調整し、読みだした撮像画像内の輝度の分布に基づいて背景をグレーに調節する。
輪郭抽出部13は、色合いが調節された撮像画像内の画像のエッジの抽出をCanny法により行いエッジ画像の生成を行う。ここで、Canny法に関しては、「Canny、J.,A computational approach to edge detection,IEEE Trans.Pattern Analysis and Machine Intellgence、8:679-714,1986」に記載されている。
histograms,IEEE Transaction on Systems,Man and Cybernetics,9(1):62-66,1979」に記載されている。
segmentation,Acoustics,Speech,and Signal
Processing,IEEE International Conference on ICASSP’82.,7:1928-1931,1982」に記載されている。
また、画像合成部16は、細胞セグメント及び背景セグメントの各々を合成し、それぞれ細胞画像と背景画像とを生成して合成画像とする。
また、細胞形態検出部19は、細胞の細胞外形(後述する細胞画像の長径と短径との比など)から細胞の成長過程を推定する。
図3は、テーブル記憶部23に書き込んで記憶される個々の細胞に関わる数値情報の種類を示す図である。
この数値情報の各種類のデータは、Name、ID、Type、OuterArea、OuterOutlineLength、OuterCenterX、OuterCenterY、OuterMaxRadius、OuterLongAxisLength、OuterShortAxisLength、OuterAxisRatio(L/S)、Round fitness、Chordiogram distance、OuterRedIntensity、OuterGreenIntensity、OuterBlueIntensity、InnerArea、InnerOutlineLength、InnerRedIntensity、InnerGreenIntensity、InnerBlueIntensityを含んでいる。以下、図3における数値情報の各々について説明する。
以下に、細胞構造抽出部18の行う数値情報の算出処理について説明する。
細胞構造抽出部18は、細胞画像内のピクセル数を計数して、このピクセル数を図3における数値情報の細胞の面積OuterAreaとして、テーブル記憶部23に書き込んで記憶させる。細胞構造抽出部18は、細胞画像の外周に配列したピクセルの数を計数して、この外周のピクセル数を図3に示す細胞の周囲の長さOuterOutlineLengthとして、テーブル記憶部23に書き込んで記憶させる。細胞構造抽出部18は、細胞画像の面積から重心を求め、撮像画像における重心のx座標及びy座標を求め、それぞれOuterCenterX、OuterCenterYとして、テーブル記憶部23に書き込んで記憶させる。このx座標は撮像画像の左端からのピクセル数であり、y座標は撮像画像の上端からのピクセル数である。
この数値情報の各種類のデータは、Name、Touch、Contaminant、Algae、Others、Cell count、Astaxanthin predictor、Astaxanthin、Chlorophyll predictor、Chlorophyllを含んでいる。以下、図4における数値情報の各々について説明する。
ここで、Touch、Contaminant、Algae、Othersについては、細胞構造抽出部18がテーブル記憶部23における細胞画像のTypeを読み込んで、それぞれの細胞数を計数して求め、求めた計数値をテーブル記憶部23の図4に示す数値情報として書き込んで記憶させる。また、Cell countについてはそれぞれの細胞数の和を求め、求めた計算値をテーブル記憶部23の図4に示す数値情報として書き込んで記憶させる。
撮像画像のすでに述べた画像解析を行い、上述した撮像画像内の数値情報を求め(色素の蓄積量以外)、細胞画像のR画素の輝度値の平均値と、B画素の輝度値の平均値との輝度比を求める。
そして、分光光度計によって、このアスタキサンチンの色素抽出液の492nm及び750nmにおける光の吸収を吸光度として測定した。この吸光度の測定結果において、アスタキサンチンの色素濃度は、以下の式により求めた。
アスタキサンチン量(μg/ml)=4.5(A492-A750)
この式において、4.5は比例定数であり、A492は492nmにおける吸光度であり、A750は750nmにおける吸光度である。このとき得られたアスタキサンチン量を、アスタキサンチンの抽出に用いた細胞数で除算して細胞当たりのアスタキサンチン量を算出した。
そして、細胞画像内のR画素及びB画素の輝度値の平均値を求め、この平均値の比である輝度比とアスタキサンチンの抽出量との対応を示す回帰式を、すなわちアスタキサンチンの量を推定する回帰式を図5の相関を求める回帰分析により求めた。
撮像画像のすでに述べた画像解析を行い、上述した撮像画像内の数値情報を求め(色素の蓄積量以外)、細胞画像のG画素の輝度値の平均値と、B画素の輝度値の平均値との輝度比を求める。
そして、分光光度計によって、このクロロフィルの色素抽出液の649nm、665nm及び750nmにおける光の吸収を吸光度として測定した。この吸光度の測定結果において、クロロフィルの色素濃度は、以下の式により求めた。
クロロフィル量(μg/ml)=
14.85(A665-A750)-5.14(A649-A750)
この式において、14.85及び5.14は比例定数であり、A665は665nmにおける吸光度であり、A649は649nmにおける吸光度であり、A750は750nmにおける吸光度である。このとき得られたクロロフィル量を、クロロフィルの抽出に用いた細胞数で除算して細胞当たりのクロロフィル量を算出した。
そして、細胞画像内のG画素及びB画素の輝度値の平均値を求め、この平均値の比である輝度比とクロロフィルの抽出量との対応を示す回帰式を、すなわちクロロフィルの量を推定する回帰式を図6の相関を求める回帰分析により求めた。
ヘマトコッカスの新鮮培地への接種後の細胞形態の経時変化を観察するため、以下のようにヘマトコッカスの細胞培養を行った。すなわち、ヘマトコッカスの撮像画像の顕微鏡撮影及び回帰式を求めるためのアスタキサンチン量の定量のための試料は以下のように調製した。なお、培地におけるヘマトコッカスの細胞培養は、連続光の下で3ヶ月間にわたって行った。
そして、Haematococcus pluvialis K0084株の培養液10mlを、300ml容量のフラスコに挿入した90mlの培地に対して接種した。また、周囲温度25℃において45μE(アインシュタイン)m-2s-1の光量子束密度の白色光の下で静置することで培養を行った。
そして、培養開始直後、7日目、14日目に30μlを、フラスコから分取して、顕微鏡観察及び撮像画像の撮像を行った。
1mm厚スライドグラスの表面に対し、ネイルエナメルにより各辺が1cmの正四角形の枠を形成し、この枠内に対してヘマトコッカスの培養液の30μlを滴下した。そして、0.17mm厚カバーグラスを用い、ネイルエナメルの枠を、ネイルエナメルにより封じてプレパラートを調製した。また、顕微鏡観察は、40倍対物レンズを装着した正立型顕微鏡を用い、顕微鏡画像の撮影にはカラーCCD(Charge Coupled Device)カメラにおいて解像度2040×1536ピクセル(画素数約300万画素)を用いた。そして、撮像した撮像画像は、RGB形式(R画素、G画素及びB画素)からなるJPEG(Joint Photographic Experts Group)画像として、画像記憶部21に対し、制御部11を介して保存した。
次に、本実施形態による細胞観察装置1の細胞観察の動作を図7及び図8を用いて説明する。図7は、本実施形態による細胞観察装置1における細胞観察の処理の動作例を示すフローチャートである。図8は、エッジ画像と色素体画像とを重ね合わせた、細胞画像の最外縁部を示す図である。図8の(a)は撮像装置100が撮像した明視野画像である撮像画像を示している。本実施形態においては、細胞に対して蛍光染色などの前処理は一切行っていない。図8の(b)は、色素体領域抽出部14の生成する色素体画像を示している。
図8の(c)は、輪郭抽出部17の抽出した細胞領域画像を示している。図8の(d)は、色素体画像と細胞領域画像とを重ね合わせた画像であり、細胞画像の最外縁部を示している。
観察者は、画像記憶部21に記憶された撮像画像における細胞画像の色合い調整の処理を、図示されていない入力手段(例えば、キーボード)から制御信号を入力することにより、細胞観察装置1に対して実行させる。
制御部11は、解析処理の制御信号が入力手段から供給されると、画像記憶部21から撮像画像のデータ(図8の(a))を読み出し、読み出した撮像画像を色合い調節部12に対して供給する。
色合い調節部12は、撮像画像のR画素、G画素及びB画素の輝度値のヒストグラムをそれぞれ集計し、最頻値の輝度を背景の輝度値として選択する。
そして、制御部11は、R画素、G画素及びB画素が調整された撮像画像を画像表示部24及び輪郭抽出部13に対して供給し、画像記憶部21に対して書き込む。以下のステップS2からの画像処理は、背景領域がグレーに調整された後の撮像画像を用いて行う。
輪郭抽出部13は、撮像画像の1ピクセルを構成するRGBの各画素のデータから、R画素のデータを抽出する。
そして、輪郭抽出部13は、抽出したR画素のデータに対して輝度の最小値と最大値が0と255となるように輝度値を調節する。
また、輪郭抽出部13は、抽出したエッジの画像を画像分割部15に対して出力する。
輪郭抽出部13におけるエッジ抽出が終了すると、制御部11は、輪郭抽出部13に対して供給した撮像画像を、色素体領域抽出部14に対して供給する。
色素体領域抽出部14は、撮像画像の1ピクセルを構成するRGBの各画素のデータから、B画素のデータを抽出する。
そして、色素体領域抽出部14は、撮像画像内において、各B画素の輝度値をOtsu法により二値化して、細胞内の色素体(アスタキサンチン及びクロロフィルの色素体)の領域である色素体領域のピクセルを抽出する。また、色素体領域抽出部14は、二値化により黒となった色素体の領域を色素体画像(図8の(b))として、画像分割部15に対して出力する。
画像分割部15は、輪郭抽出部13から供給されたエッジ画像と、色素体領域抽出部14から供給された色素体画像とを重ね合わせて、エッジ画像における余分なエッジ部分の削除と、エッジ画像に不足する細胞の形状の補足を行い、新たなエッジ画像を生成する。
次に、画像分割部15は、抽出する対象物である撮像画像における細胞の画像の領域を抽出するため、抽出されたエッジ画像の情報を用いて、water-shed法により撮像画像のセグメンテーション(領域分割)を行う。
また、画像分割部15は、セグメンテーションした領域の境界線を細線化、すなわち撮像画像におけるセグメントの境界線(ピクセル1個分の幅の境界線)を生成する。
画像合成部16は、撮像画像における1ピクセルを構成するRGBの画素、すなわちR画素、G画素及びB画素の輝度値の二値化を、上述したOtsu法により行う。
そして、画像合成部16は、二値化されて1ピクセルを構成するRGBの画素のうち少なくとも一つの輝度値が閾値以下の画素の領域である画素集合領域を抽出する。実際には、この画素集合領域が撮像画像内における細胞の画像の領域内に含まれている。
画像合成部16は、撮像画像における画素集合領域の画像に対して、セグメントの境界線を重ね合わせて、セグメント毎の上述した画素集合領域の割合を求める。そして、画像合成部16は、セグメント内における画素集合領域の割合が予め設定した割合を超えるセグメントを、撮像画像内の細胞の画像の領域である細胞セグメントとして選択する。この時点では、まだ実際の細胞の領域ではない背景領域が含まれている可能性がある。
このため、画像合成部16は、選択された細胞セグメント毎に、細胞セグメント内のR画素、G画素及びB画素各々の画素値の平均値及び分散を算出する。
そして、画像合成部16は、R画素、G画素及びB画素の少なくともいずれか一つの種類の画素の輝度値の分散の大きい細胞セグメントを、撮像画像内の細胞の撮像された領域に対応する細胞セグメントとして選択する。
セグメントの選択処理が終了すると、画像合成部16は、細胞セグメント及び背景セグメントを合成し、細胞画像と背景画像とを形成して合成画像とする。
上述した細胞セグメントと背景セグメントとの分類処理は、背景セグメントに比較して細胞セグメントにおける輝度値の分散が大きい事実を利用して行っている。
次に、細胞領域抽出部17は、合成画像における細胞画像とされる画像領域を選択するため、すでに説明したwater-shed法により、合成画像から細胞画像とされる画像領域をセグメント領域として分割する。
そして、細胞領域抽出部17は、分割されたセグメント各々のうち撮像画像の縁に接しているセグメントをTouchとして振り分ける。さらに、残りのセグメントを、Chordiogramを用いて、円状のモデル画像に近い円状セグメントと、円状のモデル画像とは異なる非円状セグメントとに振り分ける。この円に近い円状セグメントが、微細藻類の細胞、すなわちヘマトコッカスとして認識する。
また、細胞領域抽出部17は、円状セグメント以外の非円状セグメントを統合し、最終的な細胞領域画像(図8の(d))として、細胞構造抽出部に対して出力する。
細胞構造抽出部18は、すでに説明した撮像画像内における細胞画像における図3に示す数値情報を算出して、細胞画像毎にテーブル記憶部23に書き込んで記憶させる。
また、このとき、細胞形態検出部19は、細胞画像において、色素体領域抽出部14が色素体領域として検出した領域におけるR画素及びG画素の輝度値の抽出を行う。そして、細胞形態検出部19は、双方の輝度値がともに予め設定された閾値未満の場合、この細胞をツボカビと判定し、この細胞の図3の数値情報におけるTypeに対して「混入他種生物」を記入する。
制御部11は、画像記憶部21内の全ての撮像画像に対して、画像解析処理が終了したか否かの判定を行う。
このとき、制御部11は、画像記憶部21内の全ての撮像画像の画像解析処理が終了した場合、処理をステップS12に対して進める。
一方、制御部11は、画像記憶部21内の全ての撮像画像の画像解析処理が終了していない場合、処理をステップS1に戻し、新たな撮像画像を画像記憶部21から読み出し、画像解析処理を継続する。
細胞構造抽出部18は、テーブル記憶部23に記憶されている図3に示す数値テーブルから数値情報を読み出し、図4に示す数値情報Contaminant、Algae、Other、Cell countの数値情報を算出し、観察対象の細胞のフォルダ毎に書き込んで記憶させる。このフォルダには、例えば、所定の日時において、アスタキサンチン及びクロロフィルの色素量の算出に用いた複数の撮像画像における細胞画像から求めた数値情報が記憶される。
制御部11は、撮像画像の画像解析処理が終了した後、色素値抽出部20に対して、色素量の抽出処理を行わせる。
次に、色素値抽出部20は、全細胞画像のOuterRedIntensityを加算してOuterRedIntensity加算値を求め、同様に全細胞画像のOuterBlueIntensityを加算してOuterBlueIntensity加算値を求める。
次に、色素値抽出部20は、記憶部22からアスタキサンチン算出用の回帰式を読み出し、求めたAstaxanthin predictorをこの回帰式に代入し、細胞当たりのアスタキサンチンの蓄積量を求め、Astaxanthinとして、テーブル記憶部23の図4の数値情報として書き込んで記憶させる。
次に、色素値抽出部20は、全細胞画像のOuterGreenIntensityを加算してOuterGreenIntensity加算値を求め、同様に全細胞画像のOuterBlueIntensityを加算してOuterBlueIntensity加算値を求める。
次に、色素値抽出部20は、記憶部22からクロロフィル算出用の回帰式を読み出し、求めたChlorophyll predictorをこの回帰式に代入し、細胞当たりのアスタキサンチンの蓄積量を求め、Chlorophyllとして、テーブル記憶部23の図4の数値情報として書き込んで記憶させる。
図9から判るように、培地に対するヘマトコッカスの接種後に、接種直後(0週)、1週、2週後において、アスタキサンチン及びクロロフィルの蓄積量が増加している。
この結果は、実際のヘマトコッカスの細胞において、アスタキサンチンの蓄積量は1週間後にプラトーに達し、クロロフィルの蓄積量は時間経過により増加するとの知見と一致している。したがって、本実施形態による細胞観察装置1は実際に細胞を破壊して色素体の蓄積量を測定せずとも、培養している状態のまま色素の蓄積量を正確に推定することが可能である。
図10は、細胞観察装置1が画像解析から求めた細胞画像の数値情報(OuterAxisRatio(L/S)、OuterArea)と、目視による細胞の分類との対応を示す図である。図10において、縦軸は細胞画像の長径及び短径の比OuterAxisRatio(L/S)を示し、横軸は細胞画像の面積OuterAreaを示している。
また、「+」マークは目視により遊走子として判定された細胞であり、「〇」マークが目視により遊走子からパルメラ細胞へ移行途中の移行細胞として判定された細胞であり、「×」マークが目視によりパルメラ細胞(アスタキサンチンを色素として産生するヘマトコッカスの細胞)として判定された細胞である。
新鮮培地に移した後のヘマトコッカスの細胞に対して、20枚の撮像画像においてヘマトコッカスの細胞を目視で遊走子とパルメラ細胞と判別できない移行状態の細胞との3種類に分類した。
そして、撮像画像を本実施形態の細胞観察装置1により解析し、ヘマトコッカスの細胞と判定した計138種類の細胞についての定量値を利用し、遊走子とパルメラ細胞との判別を行った。細胞のサイズ(面積)と細胞の長軸短軸比について、上述したように遊走子を「+」で表し、パルメラ細胞を「×」で表し、移行状態の細胞を「○」で表してプロットされている。この図10のグラフにおいて、細胞の面積が3000ピクセル未満の領域に遊走子の87%が、細胞の面積が3000ピクセル以上の領域にパルメラ細胞の96%が含まれることが判明した。
次に、細胞画像内のピクセルにおける赤チャンネル(R画素)、緑チャンネル(G画素)及び青チャンネル(B画素)の輝度値を用いたアスタキサンチン(カロテノイド)量及びクロロフィル量の各々の推定について説明する。アスタキサンチンは、カロテノイドの一種である。
すでに、図5においては、細胞画像内におけるR画素とB画素との輝度値の平均値を用いた回帰式により、アスタキサンチン量の推定の説明をしている。また、図6においては、細胞画像内におけるG画素とB画素との各々の輝度値を用いた回帰式により、クロロフィル量の推定の説明をしている。
しかしながら、本実施形態においては上述したように、細胞画像内におけるR画素及びB画素の輝度値の平均値に対して、細胞画像内におけるG画素の輝度値の平均値を加え、細胞画像内におけるR画素、G画素及びB画素の各々の輝度値の平均値から、細胞の産生する色素量を求める重回帰式を用いて、アスタキサンチン量及びクロロフィル量の各々を求めている。アスタキサンチン量を求める重回帰式とクロロフィル量を求める重回帰式とは、各々異なる重回帰式(後述)であり、予め記憶部22に書き込まれて記憶されている。
Ychl=-0.46×IR+0.56×IG-0.83×IB-72.01
また、以下に示すカロテノイド濃度(pg/cell)の重回帰式(図15にも提示)は、1個の細胞あたりに含まれるカロテノイド量を推定する式である。
Ycar=0.75×IR-0.22×IG-0.27×IB-61.83
上述した重回帰式は、以下の基本式における係数a、b、c、dの各々の値を重回帰分析で求めて生成した。
Y=aIR+bIG+cIB+d
上記式において、係数aは、色素体のピクセルの赤チャンネルの輝度値の平均値に乗算する係数である。係数bは、色素体のピクセルの緑チャンネルの輝度値の平均値に乗算する係数である。係数cは、色素体のピクセルの青チャンネルの輝度値の平均値に乗算する係数である。係数dは、重回帰式における定数である。
次に、ランダムフォレスト(Random forests)法によって、観察対象の培養細胞が遊走子とパルメロイドとのいずれかであるかの判別を行う判別処理について説明する。本項において、基本的には、ランダムフォレスト法による機械学習を用いて、撮像画像に含まれる細胞画像が示す細胞各々を、遊走子とパルメロイドとにそれぞれ分離するクラスタリング処理用の樹木モデルを生成する。そして、生成された樹木モデルを記憶部22に予め書き込んで記憶させておく。本実施形態においては、パルメロイドはパルメラ細胞とも言う。
また、細胞形態検出部19は、クラスタリング結果の正解を入力されることにより、ランダムフォレスト法による機械学習を繰り返し、樹木モデルを更新していくように構成しても良い。
観察者が顕微鏡で観察した場合、パルメロイドと遊走子との区別は遊走子に鞭毛などがあって容易につくが、一方、培養している大量の細胞の全てを人間が判別し、パルメロイドと遊走子との数の比などを検出することは困難である。
上述した点を改善するため、樹木モデルによる培養細胞各々の数値的なクラスタリングのみでなく、クラスタリングの結果を表示装置にビジュアル的に表示することにより、遊走子とパルメロイドとの比率を明確に通知することができる。
また、トレーニングデータ及びテストデータの各々の細胞は、新鮮培地に移した2日目あるいは3日目のヘマトコッカスの細胞を母集団として、この母集団からから抽出した682個の細胞を341個ずつに分けて、一方をトレーニングデータの細胞、他方をテストデータの細胞としたものである。
また、図17から図20の各々は、細胞培養を開始してから2日目あるいは3日目の培養後に判別を行っている。したがって、細胞培養を初めてから、時系列に上述した樹木モデルにより判別を行った場合、クラスタリング結果により、遊走子とパルメロイドとの比率の変化などが図17から図20に示すように、ビジュアル的に確認することができる。
さらに、トレーニングデータを異なる細胞培養の株を用いて増やし、樹木モデルを更新することにより、より遊走子とパルメロイドとのクラスタリングの精度を向上させることができ、細胞の培養の経過観察が撮像した細胞画像により容易にできるようになる。
上述したランダムフォレスト法における樹木モデルの生成に、図21から図45までに示した25個のパラメータが用いられている。
次に、図21から図45までの25個を含む細胞形態検出部19が抽出するパラメータから、ヘマトコッカスの遊走子からパルメロイドへの成長過程における細胞形態の経時変化及び条件依存を検出する評価用のパラメータを主成分分析により抽出した。
この主成分分析において、第1主成分PC1の寄与率は51%であった。第1主成分PC1は、上述した図21から図45までの25個を含む細胞構造抽出部18が抽出するパラメータの内、長軸の長さOuterLongAxisLength、細胞の面積OuterAreaを含む8個のパラメータ(後述)と高い相関があることが解った。
この8個のパラメータとしては、例えば、図21に示す細胞の周囲の長さOuterOutlineLength及び図22に示す細胞画像の重心から外周までの最大幅OuterMaxRadiusなどは、時間変化に対して線形的な変化を有し、第1主成分PC1と高い相関を有している。
この図46から解るように第1主成分の値は時間と共に減少しているが、LL条件とLD条件とは変化の状態が異なる。しかしながら、LL条件及びLD条件の各々により培養した細胞ともに、第1主成分PC1との相関があり、当然に時間変化に対しても相関がある。このため、LL条件及びLD条件の双方に対して、第1主成分PC1と相関のあるパラメータは、観察対象細胞の時間経過における成長過程での形態変化を検出するパラメータとして用いることができることが解る。
また、それ以降の時間経過における鞭毛の消失の時間的変化に対し、LL条件及びLD条件の各々の培養の環境下の細胞間に差はないことが解る。このパラメータから、培養を開始してから鞭毛を発生させるまでの成長の速度は、LL条件下での環境による培養の方が、LD条件下での環境による培養より速いことが解る。しかしながら、鞭毛が発生した後、遊走子からパルメロイドに成長するまでの過程は、LL条件及びLD条件の各々の環境下による培養の差がないことも解る。
一方、LD条件の環境下の培養においては、細胞を撮像した細胞画像におけるOuterMeanGreenIntensityの数値が大きく低下している。これは、OuterMeanGreenIntensityが、LD条件の環境下において培養した細胞と、LL条件下で培養した細胞とを分離することができるパラメータであることが解る。
したがって、第1主成分PC1と相関が少ないパラメータから、培養の条件及び環境に依存するパラメータを抽出することにより、上述したように、LL条件下とLD条件下との各々において成長した細胞のクラスタリングを可能とする。
上述したように、本実施形態による細胞観察装置が抽出したパラメータから、観察対象細胞の経時的変化と条件及び環境依存的変化との各々の特徴を有するパラメータを主成分分析で抽出できた。
また、主成分分析で抽出したパラメータにより、観察対象細胞の成長における経時変化のダイナミズムを評価することができる。
次に、判別分析により、クロレラ細胞の異なる株の判別分析を行った。例えば、本実施形態においては、クロレラの野生(WT)株と、この野生株から派生した変異(ミュータント)株PkE6と、変異株PkE8との判別分析を行った。判別分析は、事前に与えられているデータの各々が異なるグループに分離されることが明らかな場合、新しいデータが得られた際に、いずれのグループに入るかを判別する(クラスタリングを行う)ための基準(判別関数)を得るための手法である。以下の処理は、細胞形態検出部19が図21から図45に示すパラメータのデータを用いて行う。
この図50から解るように、粒子径については、変異株PkE6が大きいことが、本実施形態である細胞観察装置1による撮像画像における細胞画像の各々に対する画像解析でも確認できた。
その結果、変異株PkE6が最も大きく、次に変異株PkE8が大きく、野生株(WT)が最も小さいことが解る。また、図50から変異株PkE8にも変異株PkE6と同等の大きさの細胞の集団が存在していることが解る。
上述した結果から、図51に示すパラメータを用いて判別分析を行うことにより、野生(WT)株、変異株PkE8、変異株PkE6の各々を判別することが可能であることが予想される。
上述したように、判別関数LD1は、細胞画像の細胞内あるいは色素体内におけるピクセルにおける各々の色のチャンネルの輝度を示すパラメータと相関が高い。一方、判別関数LD2は、判別関数LD2は、細胞画像の細胞の大きさを示すパラメータと相関が高い。
このように、野生(WT)株、変異株PkE8、変異株PkE6の各々がそれぞれのグループとして分離(クラスタリング)されていることが図52から解る。
図54の(a)は、細胞のピクセルの赤チャンネルの輝度値の平均値と、その平均値を有する細胞の出現率との対応を示している。図54の(a)において、縦軸が細胞の出現率を示し、横軸がパラメータOuterMeanRedIntensityの数値を示している。OuterMeanRedIntensityについては、変異株PkE8の輝度値の平均値は、野生(WT)株及び変異株PkE6と比較して最も高い。変異株PkE6の輝度値の平均値は、変異株PkE8と同様に高い集団と、野生(WT)株及び変異株PkE8より低い集団との2つのピークが見られる。野生(WT)株の輝度値の平均値は、変異株PkE8より低く、変異株PkE6の輝度値の低い集団より高い。
また、「コンピュータ読み取り可能な記録媒体」とは、フレキシブルディスク、光磁気ディスク、ROM、CD-ROM等の可搬媒体、コンピュータシステムに内蔵されるハードディスク等の記憶装置のことをいう。さらに「コンピュータ読み取り可能な記録媒体」とは、インターネット等のネットワークや電話回線等の通信回線を介してプログラムを送信する場合の通信線のように、短時間の間、動的にプログラムを保持するもの、その場合のサーバやクライアントとなるコンピュータシステム内部の揮発性メモリのように、一定時間プログラムを保持しているものも含むものとする。また上記プログラムは、前述した機能の一部を実現するためのものであっても良く、さらに前述した機能をコンピュータシステムにすでに記録されているプログラムとの組み合わせで実現できるものであっても良い。
Claims (6)
- 一層の細胞を撮像した撮像画像から、当該撮像画像内の細胞の画像のエッジの画素を抽出し、抽出したエッジの画素からなるエッジ画像を生成する輪郭抽出部と、
前記撮像画像における細胞の画像の色素体の画素を抽出し、この抽出した色素体の画素からなる色素体画像を生成する色素体領域抽出部と、
前記エッジ画像と前記色素体画像とを重ね合わせて合成された合成画像において、前記撮像画像における細胞の画像領域と背景の画像領域とを、前記画素の輝度値の分散により検出し、前記撮像画像における細胞の画像領域を検出する画像合成部と
を備えることを特徴とする細胞観察装置。 - 前記合成画像において色素体の存在する細胞の画像領域を対象細胞画像とし、色素体の存在しない細胞の画像領域を非対象細胞画像として検出し、前記合成画像における全細胞の画像における当該非対象細胞画像の比率を求める細胞形態検出部
をさらに備えることを特徴とする請求項1に記載の細胞観察装置。 - 前記細胞の画像領域における色素体の輝度値から前記色素体の量である色素体量を算出する色素値抽出部
をさらに備えることを特徴とする請求項1または請求項2に記載の細胞観察装置。 - 前記細胞の画像領域から色素体の平均輝度値を求めておき、前記撮像画像を撮像した当該細胞から色素体を抽出して色素体量を求め、平均輝度値と細胞当たりの色素体量との回帰式を予め求めて記憶部に記憶させておき、
前記色素値抽出部が前記細胞画像における平均輝度値を求め、前記回帰式から細胞当たりの色素体量を求めることを特徴とする請求項3に記載の細胞観察装置。 - 輪郭抽出部が、一層の細胞を撮像した撮像画像から、当該撮像画像内の細胞の画像のエッジの画素を抽出し、抽出したエッジの画素からなるエッジ画像を生成する輪郭抽出過程と、
色素体領域抽出部が、前記撮像画像における細胞の画像の色素体の画素を抽出し、この抽出した色素体の画素からなる色素体画像を生成する色素体領域抽出過程と、
画像合成部が、前記エッジ画像と前記色素体画像とを重ね合わせて合成された合成画像において、前記撮像画像における細胞の画像領域と背景の画像領域とを、前記画素の輝度値の分散により検出し、前記撮像画像における細胞の画像領域を検出する画像合成過程と
を含むことを特徴とする細胞観察方法。 - 細胞の形状を観察する細胞観察装置の動作をコンピュータに実行させるプログラムであり、
コンピュータを、
一層の細胞を撮像した撮像画像から、当該撮像画像内の細胞の画像のエッジの画素を抽出し、抽出したエッジの画素からなるエッジ画像を生成する輪郭抽出手段、
前記撮像画像における細胞の画像の色素体の画素を抽出し、この抽出した色素体の画素からなる色素体画像を生成する色素体領域抽出手段、
前記エッジ画像と前記色素体画像とを重ね合わせて合成された合成画像において、前記撮像画像における細胞の画像領域と背景の画像領域とを、前記画素の輝度値の分散により検出し、前記撮像画像における細胞の画像領域を検出する画像合成手段
として機能させるためのプログラム。
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2013353154A AU2013353154B2 (en) | 2012-11-28 | 2013-11-27 | Cell monitoring device, cell monitoring method and program thereof. |
DK13857702.8T DK2927311T3 (da) | 2012-11-28 | 2013-11-27 | Celleovervågningsindretning, celleovervågningsfremgangsmåde og program dertil |
NZ708612A NZ708612A (en) | 2012-11-28 | 2013-11-27 | Cell monitoring device, cell monitoring method and program thereof |
JP2014549862A JP6278519B2 (ja) | 2012-11-28 | 2013-11-27 | 細胞観察装置、細胞観察方法及びそのプログラム |
EP13857702.8A EP2927311B1 (en) | 2012-11-28 | 2013-11-27 | Cell observation device, cell observation method and program thereof |
US14/647,904 US9477875B2 (en) | 2012-11-28 | 2013-11-27 | Cell monitoring device, cell monitoring method and program thereof |
IL238988A IL238988B (en) | 2012-11-28 | 2015-05-25 | Cell monitoring device, cell monitoring method and program therefor |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012259880 | 2012-11-28 | ||
JP2012-259880 | 2012-11-28 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014084255A1 true WO2014084255A1 (ja) | 2014-06-05 |
Family
ID=50827887
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/081894 WO2014084255A1 (ja) | 2012-11-28 | 2013-11-27 | 細胞観察装置、細胞観察方法及びそのプログラム |
Country Status (8)
Country | Link |
---|---|
US (1) | US9477875B2 (ja) |
EP (1) | EP2927311B1 (ja) |
JP (1) | JP6278519B2 (ja) |
AU (1) | AU2013353154B2 (ja) |
DK (1) | DK2927311T3 (ja) |
IL (1) | IL238988B (ja) |
NZ (1) | NZ708612A (ja) |
WO (1) | WO2014084255A1 (ja) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2016123366A (ja) * | 2015-01-05 | 2016-07-11 | 株式会社ニコン | 細胞評価装置、インキュベータ、細胞評価方法、プログラム、及び細胞の培養方法 |
JP2017042146A (ja) * | 2015-08-28 | 2017-03-02 | 株式会社エヌテック | 微生物の検出方法、微生物の検出装置及びプログラム |
JP2017522563A (ja) * | 2014-07-22 | 2017-08-10 | オリンパス株式会社 | 混合物の光学的な組成解析 |
JP2020047190A (ja) * | 2018-09-21 | 2020-03-26 | 株式会社Screenホールディングス | 画像処理方法、コンピュータプログラムおよび記録媒体 |
JP2020532016A (ja) * | 2017-09-12 | 2020-11-05 | 日本電気株式会社 | 画像処理装置、画像処理方法及びプログラム |
Families Citing this family (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9522396B2 (en) | 2010-12-29 | 2016-12-20 | S.D. Sight Diagnostics Ltd. | Apparatus and method for automatic detection of pathogens |
EP2798350B1 (en) | 2011-12-29 | 2021-07-28 | Sight Diagnostics Ltd. | Methods and systems for detecting a pathogen in a biological sample |
EP2973391B1 (en) * | 2013-03-15 | 2018-11-14 | Koninklijke Philips N.V. | Determining a residual mode image from a dual energy image |
EP2999988A4 (en) | 2013-05-23 | 2017-01-11 | S.D. Sight Diagnostics Ltd. | Method and system for imaging a cell sample |
IL227276A0 (en) | 2013-07-01 | 2014-03-06 | Parasight Ltd | A method and system for obtaining a monolayer of cells, for use specifically for diagnosis |
CN105659151B (zh) | 2013-08-26 | 2018-06-29 | 思迪赛特诊断有限公司 | 数码显微系统与方法 |
WO2016030897A1 (en) | 2014-08-27 | 2016-03-03 | S.D. Sight Diagnostics Ltd | System and method for calculating focus variation for a digital microscope |
US10684206B2 (en) | 2014-09-29 | 2020-06-16 | Biosurfit, S.A. | Systems and related methods of imaging and classifying cells in a blood sample |
WO2016204027A1 (ja) * | 2015-06-18 | 2016-12-22 | コニカミノルタ株式会社 | 目的生体物質の解析装置、解析システム、解析方法および解析プログラム |
JP6952683B2 (ja) | 2015-09-17 | 2021-10-20 | エス.ディー.サイト ダイアグノスティクス リミテッド | 身体試料中の実体を検出する方法および装置 |
US11733150B2 (en) | 2016-03-30 | 2023-08-22 | S.D. Sight Diagnostics Ltd. | Distinguishing between blood sample components |
US11307196B2 (en) | 2016-05-11 | 2022-04-19 | S.D. Sight Diagnostics Ltd. | Sample carrier for optical measurements |
AU2017263807B2 (en) | 2016-05-11 | 2023-02-02 | S.D. Sight Diagnostics Ltd | Performing optical measurements on a sample |
GB201610425D0 (en) | 2016-06-15 | 2016-07-27 | Q-Linea Ab | Analysis of images of biological material |
JP6900704B2 (ja) * | 2017-02-28 | 2021-07-07 | 株式会社島津製作所 | 細胞観察システム |
WO2019084513A1 (en) * | 2017-10-26 | 2019-05-02 | Essenlix Corporation | QUICK MEASUREMENT OF PLATELETS |
WO2019097387A1 (en) | 2017-11-14 | 2019-05-23 | S.D. Sight Diagnostics Ltd | Sample carrier for optical measurements |
WO2020240555A1 (en) * | 2019-05-29 | 2020-12-03 | Ramot At Tel-Aviv University Ltd. | System and method for reconstructing morphology and dynamics of biological cells from holographic images |
US11394774B2 (en) * | 2020-02-10 | 2022-07-19 | Subash Sundaresan | System and method of certification for incremental training of machine learning models at edge devices in a peer to peer network |
CN117253227B (zh) * | 2023-10-23 | 2024-06-18 | 深圳市茵冠生物科技有限公司 | 一种人工智能协同自动化设备的监控方法与系统 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007097584A (ja) | 2005-09-06 | 2007-04-19 | Yamaha Motor Co Ltd | アスタキサンチン含有量の高い緑藻およびその製造方法 |
JP2009175334A (ja) * | 2008-01-23 | 2009-08-06 | Olympus Corp | 顕微鏡システム、画像生成方法、及びプログラム |
JP2010063403A (ja) * | 2008-09-10 | 2010-03-25 | Ihi Corp | 微生物の簡易計数方法 |
JP2011030494A (ja) | 2009-07-31 | 2011-02-17 | Kirin Brewery Co Ltd | 細胞形態定量値を用いる酵母の生理状態の評価方法 |
US20120014608A1 (en) * | 2010-07-15 | 2012-01-19 | Sony Corporation | Apparatus and method for image processing |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6956961B2 (en) * | 2001-02-20 | 2005-10-18 | Cytokinetics, Inc. | Extracting shape information contained in cell images |
US7219016B2 (en) * | 2001-04-20 | 2007-05-15 | Yale University | Systems and methods for automated analysis of cells and tissues |
US7466872B2 (en) | 2005-06-20 | 2008-12-16 | Drvision Technologies Llc | Object based boundary refinement method |
GB2433985A (en) * | 2006-01-09 | 2007-07-11 | Cytokinetics Inc | Characterization of features within the boundary regions of biological cells |
EP2430448A1 (en) * | 2009-05-15 | 2012-03-21 | SRU Biosystems, Inc. | Detection of changes in cell populations and mixed cell populations |
WO2012035504A1 (en) * | 2010-09-14 | 2012-03-22 | Ramot At Tel-Aviv University Ltd. | Cell occupancy measurement |
-
2013
- 2013-11-27 JP JP2014549862A patent/JP6278519B2/ja active Active
- 2013-11-27 EP EP13857702.8A patent/EP2927311B1/en not_active Not-in-force
- 2013-11-27 WO PCT/JP2013/081894 patent/WO2014084255A1/ja active Application Filing
- 2013-11-27 NZ NZ708612A patent/NZ708612A/en not_active IP Right Cessation
- 2013-11-27 AU AU2013353154A patent/AU2013353154B2/en not_active Ceased
- 2013-11-27 DK DK13857702.8T patent/DK2927311T3/da active
- 2013-11-27 US US14/647,904 patent/US9477875B2/en active Active
-
2015
- 2015-05-25 IL IL238988A patent/IL238988B/en active IP Right Grant
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007097584A (ja) | 2005-09-06 | 2007-04-19 | Yamaha Motor Co Ltd | アスタキサンチン含有量の高い緑藻およびその製造方法 |
JP2009175334A (ja) * | 2008-01-23 | 2009-08-06 | Olympus Corp | 顕微鏡システム、画像生成方法、及びプログラム |
US20090213214A1 (en) * | 2008-01-23 | 2009-08-27 | Tatsuki Yamada | Microscope System, Image Generating Method, and Program for Practising the Same |
JP2010063403A (ja) * | 2008-09-10 | 2010-03-25 | Ihi Corp | 微生物の簡易計数方法 |
JP2011030494A (ja) | 2009-07-31 | 2011-02-17 | Kirin Brewery Co Ltd | 細胞形態定量値を用いる酵母の生理状態の評価方法 |
US20120014608A1 (en) * | 2010-07-15 | 2012-01-19 | Sony Corporation | Apparatus and method for image processing |
JP2012021904A (ja) * | 2010-07-15 | 2012-02-02 | Sony Corp | 画像処理装置および方法 |
CN102375992A (zh) * | 2010-07-15 | 2012-03-14 | 索尼公司 | 图像处理装置和方法 |
Non-Patent Citations (15)
Title |
---|
"E. B. C. Analytica Microbiologica.", J. INST. BREW., vol. 83, 1977, pages 109 - 118 |
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, vol. 75, 2009, pages 2545 - 2553 |
BEUCHER, S.: "Watershed of functions and picture segmentation", ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, IEEE INTERNATIONAL CONFERENCE ON ICASSP'82, vol. 7, 1982, pages 1928 - 1931 |
CANNY, J.: "A computational approach to edge detection", IEEE TRANS. PATTERN ANALYSIS AND MACHINE INTELLGENCE, vol. 8, 1986, pages 679 - 714 |
COLLINS AM . ET AL.: "CAROTENOID DISTRIBUTION IN LIVING CELLS OF HAEMATOCOCCUS PLUVIALIS (CHLOROPHYCEAE)", PLOS ONE, vol. 6, no. 9, 2011, pages E24302, XP055255386 * |
FUNGAL BIOLOGY, vol. 115, 2011, pages 803 - 811 |
JOURNAL OF PHYCOLOGY, vol. 30, 1994, pages 829 - 833 |
KAMATH SB. ET AL.: "Digital image processing-an alternate tool for monitoring of pigment levels in cultured cells with special reference to green alga Haematococcus pluvialis", BIOSENS. BIOELECTRON., vol. 21, no. 5, November 2005 (2005-11-01), pages 768 - 773, XP005122222 * |
MYCOLOGICAL RESEARCH, vol. 112, 2008, pages 70 - 81 |
OHNUKI S. ET AL.: "Image-based monitoring system for green algal Haematococcus pluvialis (Chlorophyceae) cells during culture", PLANT CELL PHYSIOL., vol. 54, no. 11, 2013, pages 1917 - 1929, XP055194636 * |
OTSU N: "A threshold selection method from gray-level histograms", IEEE TRANSACTION ON SYSTEMS, MAN AND CYBERNETICS, vol. 9, no. 1, 1979, pages 62 - 66 |
SATORU NOGAMI ET AL.: "HaematoCalMorph: Haematococcus pluvialis no Temeno Takino Saibo Gazo Shori Software", THE BULLETIN OF JAPANESE SOCIETY OF PHYCOLOGY, vol. 61, no. 1, 2013, pages 44, B17, XP008179264 * |
See also references of EP2927311A4 |
SU CH ET AL.: "SIMULTANEOUS ESTIMATION OF CHLOROPHYLL A AND LIPID CONTENTS IN microalgae by three-color analysis", BIOTECHNOL. BIOENG., vol. 99, no. 4, 2008, pages 1034 - 1039, XP055255384 * |
TOSHEV, A.; TASKAR, B.; DANIILIDIS, K.: "Object detection via boundary structure segmentation", COMPUTER VISION AND PATTERN RECOGNITION, 2010, pages 950 - 957 |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2017522563A (ja) * | 2014-07-22 | 2017-08-10 | オリンパス株式会社 | 混合物の光学的な組成解析 |
JP2016123366A (ja) * | 2015-01-05 | 2016-07-11 | 株式会社ニコン | 細胞評価装置、インキュベータ、細胞評価方法、プログラム、及び細胞の培養方法 |
JP2017042146A (ja) * | 2015-08-28 | 2017-03-02 | 株式会社エヌテック | 微生物の検出方法、微生物の検出装置及びプログラム |
JP2020532016A (ja) * | 2017-09-12 | 2020-11-05 | 日本電気株式会社 | 画像処理装置、画像処理方法及びプログラム |
US11195299B2 (en) | 2017-09-12 | 2021-12-07 | Nec Corporation | Image processing device, image processing method and storage medium |
JP2020047190A (ja) * | 2018-09-21 | 2020-03-26 | 株式会社Screenホールディングス | 画像処理方法、コンピュータプログラムおよび記録媒体 |
JP7099923B2 (ja) | 2018-09-21 | 2022-07-12 | 株式会社Screenホールディングス | 画像処理方法、コンピュータプログラムおよび記録媒体 |
Also Published As
Publication number | Publication date |
---|---|
IL238988B (en) | 2020-11-30 |
JPWO2014084255A1 (ja) | 2017-01-05 |
EP2927311A1 (en) | 2015-10-07 |
IL238988A0 (en) | 2015-07-30 |
DK2927311T3 (da) | 2020-01-13 |
US20150302237A1 (en) | 2015-10-22 |
JP6278519B2 (ja) | 2018-02-21 |
AU2013353154A1 (en) | 2015-06-18 |
US9477875B2 (en) | 2016-10-25 |
AU2013353154B2 (en) | 2016-06-23 |
NZ708612A (en) | 2016-04-29 |
EP2927311A4 (en) | 2016-08-17 |
EP2927311B1 (en) | 2019-11-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6278519B2 (ja) | 細胞観察装置、細胞観察方法及びそのプログラム | |
Zhang et al. | A comprehensive review of image analysis methods for microorganism counting: from classical image processing to deep learning approaches | |
Hung et al. | Applying faster R-CNN for object detection on malaria images | |
CN101701916B (zh) | 一种玉米品种快速鉴定、鉴别方法 | |
CN102288606B (zh) | 基于机器视觉的花粉活力测定方法 | |
Dörge et al. | Direct identification of pure Penicillium species using image analysis | |
US20130109051A1 (en) | Method for detecting microorganisms, device for detecting microorganisms and program | |
CN107580715A (zh) | 用于自动计数微生物菌落的方法和系统 | |
CN104651462A (zh) | 基于显微图像分析的稻瘟病菌孢子检测方法 | |
Capus et al. | Application of high-content image analysis for quantitatively estimating lipid accumulation in oleaginous yeasts with potential for use in biodiesel production | |
Timoumi et al. | Dynamic behavior of Yarrowia lipolytica in response to pH perturbations: dependence of the stress response on the culture mode | |
Coltelli et al. | Automatic and real time recognition of microalgae by means of pigment signature and shape | |
WO2021095381A1 (ja) | 評価対象の評価方法、画像処理装置及び評価対象の評価システム | |
Ogawa et al. | Noise-free accurate count of microbial colonies by time-lapse shadow image analysis | |
Zhang et al. | An automated bacterial colony counting system | |
CN109948544B (zh) | 一种目标菌落自动定位与识别方法 | |
Ebrahimi et al. | Quantitative determination of mold growth and inhibition by multispectral imaging | |
Clemmensen et al. | A method for comparison of growth media in objective identification of Penicillium based on multi-spectral imaging | |
Wit et al. | Application of an artificial neural network in the enumeration of yeasts and bacteria adhering to solid substrata | |
Katsuragi et al. | Single‐Cell Sorting of Microorganisms by Flow or Slide‐Based (Including Laser Scanning) Cytometry | |
Fischer et al. | Quality assurance and good laboratory practice in the mycological laboratory–compilation of basic techniques for the identification of fungi | |
Kaur et al. | A novel methodology for automatic bacterial colony counter | |
Leroux et al. | Hyperspectral imaging applied to microbial categorization in an automated microbiology workflow | |
US20240193968A1 (en) | Analyzing microscope images of microalgae culture samples | |
Robinson et al. | Morphological characterisation of yeast colony growth on solid media using image processing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13857702 Country of ref document: EP Kind code of ref document: A1 |
|
DPE1 | Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101) | ||
ENP | Entry into the national phase |
Ref document number: 2014549862 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 238988 Country of ref document: IL |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2013857702 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14647904 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2013353154 Country of ref document: AU Date of ref document: 20131127 Kind code of ref document: A |