JP2001174456A - Device and method for subclassification of leukocyte - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device for automating all or one portion of work related to concrete constituent analysis in a liquid to be examined and subclassifying leukocyte automatically and accurately using image processing and recognition techniques for a specimen where the leukocyte has appeared in the liquid to be examined. SOLUTION: The leukocyte subclassification device is provided with an image pickup stage for picking up the image of liquid to be examined on a light transmissible plate, a means for enlarging a sample image included in the liquid to be examined, a means for picking up the image of the sample, and a means for identifying the leukocyte constituent being subclassified by processing the image that has been picked up.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for finely classifying leukocytes to which image processing and recognition techniques by a computer are applied.

[0002]

2. Description of the Related Art Conventionally, there is a test called urine sediment test as one of the methods for diagnosing renal and urinary tract diseases. In urine sediment tests, centrifuge the urine sample, use an aspirator or pipette, or remove the supernatant by decantation.Apply a certain amount of the remaining residue components to a slide glass, place a cover glass on the slide glass and attach the sample. Each stage is set on a microscope and classified and analyzed for formed components (red blood cells, white blood cells, epithelial cells, casts, microorganisms, crystal salts, etc.). By analysis of these components in urine, information on the pathological conditions of the kidney and urinary system can be grasped sharply and accurately, and renal diseases can be distinguished from urinary tract diseases.
The presence and activity of renal and urinary tract lesions can be evaluated, and treatment can be evaluated by monitoring the progress.

[0003] Among the urine particles, leukocytes are hardly contained in urine of healthy individuals. If a large number of leukocytes are found in the urine, a urinary tract infection such as renal pelvic nephritis, cystitis, prostatitis, or urethritis is suspected. Leukocytes found in urine are mainly neutrophils. Sometimes eosinophils, lymphocytes and monocytes are mixed. Usually, neutrophils in acute inflammation, lymphocytes and monocytes in chronic inflammation, and eosinophils in allergic diseases are observed.

Neutrophils are stained with Sternheimer-Malbin stain (SM
Staining) and Sternheiner (S-staining), swelling among dark-stained cells (dark cells), pale blue to colorless pale-stained cells, and pale-stained cells, depending on the staining form There are cases where the cells are classified into three types: glitter cells in which granules that carry out Brownian motion in the cytoplasm are recognized.
Generally, the gray cells are considered to be senescent or dead leukocytes, and the bright cells and lightly-stained cells are considered to be active and live leukocytes.
If neutrophils are light-stained cells and a large number of bacteria and transitional epithelial cells are found at the same time, the possibility of cystitis is high, and if neutrophils are bright cells and the patient is accompanied by fever, pelvic nephritis is suspected. In the case of female urine, if neutrophils are densely stained cells and squamous epithelial cells and bacteria are simultaneously observed, contamination with vaginal secretions is considered. In this case, there are many mergers of Trichomonas. Eosinophils occasionally appear in large numbers during allergic cystitis and parasitic diseases seen in drug administration and are sometimes found together with rhombic Charco-Leiden crystals. Lymphocytes may be present in the urinary tract with lymphatic thinning and chyleuria or during rejection after kidney transplantation. Monocytes are found during recovery from urinary tract infections.

[0005] In the urinary sediment examination as described above, when leukocytes in urine are recognized, more detailed information on the disease state can be obtained by further finely classifying and analyzing the leukocytes, which is of great clinical significance. On the other hand, the classification of leukocytes by microscopic observation requires a skillful technique, and there is naturally an individual difference in judgment, and there is a problem in accuracy. In addition, all the work processes of urine sediment inspection are performed manually, which places a heavy burden on laboratory technicians.

[0006] Also in blood tests, subclassification of leukocytes has clinical significance. Leukocytes in peripheral blood of healthy individuals include lymphocytes, monocytes, neutrophils, eosinophils, and basophils. These have different functions, and can contribute to the diagnosis of disease by counting leukocytes in blood by type. For example, an increase in neutrophils is seen in inflammation, myocardial infarction, leukemia, etc.
It is seen in viral diseases, aplastic anemia, agranulocytosis, etc. Increases in monocytes are seen in convalescent phases of infection, monocytic leukemia, and the like.

[0007] One of the most commonly used methods for classifying and counting leukocytes in blood has been called a blood image appraisal (calculation method). In this method, blood is smeared on a slide glass, blood cells are fixed, stained, observed under a microscope, and the morphological characteristics of each leukocyte (white blood cell size, nuclear morphology, cytoplasmic morphology) , The presence or absence of granules), the degree of staining, and the like, the laboratory technician determines which blood cells are, and classifies and counts them. At this time, in a general laboratory, 100 to 200 white blood cells are counted, and the percentage (%) of each blood cell in the total number of white blood cells is used as a measured value. However, this method requires complicated sample preparation operations such as smearing, fixation, and staining of blood before microscopic observation, and classification and counting using a microscope can distinguish slight differences in blood cells. The work that has to be done places a heavy burden on engineers. In addition, the number of white blood cells to be counted is small, and the blood cells on the smear sample may have an uneven distribution, so that it is difficult even for a skilled laboratory technician to obtain a highly reproducible measurement value.

As described above, there is a strong demand for a method capable of automatically classifying and counting leukocytes in urine and blood tests. In recent years, as a device for automating these inspection tasks, a stain solution is mixed with a test solution without preparing a smear of a specimen, and then flown into a flow cell in a suspended state, using a physical statistical method or an optical method. For example, there is a method of performing an automatic analysis (Japanese Patent Laid-Open No. 4-337460,
JP-A-5-296915, JP-A-5-32288
No. 5).

[0009] However, the conventional automation technology does not automatically classify the leukocyte components in urine automatically. For the leukocyte subclassification in blood, each leukocyte is classified into several items in the same manner as a laboratory technician. It was difficult to classify quickly, and there was a problem in the analysis accuracy.

[0010]

DISCLOSURE OF THE INVENTION The present invention automates all or a part of the work relating to the analysis of solid components in a test liquid,
For a sample in which leukocytes appeared in the test solution, image processing,
An object of the present invention is to provide an apparatus for automatically and precisely classifying white blood cells by using recognition technology.

[0011]

Means for Solving the Problems As a result of intensive studies, the present inventors have automatically taken a still image of the leukocyte component in a test solution and analyzed the taken image with an analysis circuit having a high degree of discrimination ability. By processing, the present invention has been completed.

That is, the present invention has the following configuration. (1) An imaging stage for imaging a test solution on a light-transmitting plate or in a flow cell, a unit for enlarging a sample image contained in a test solution, a unit for imaging the sample image, and imaging Means for processing the processed image to identify the finely classified white blood cell component. (2) The leukocyte fine classification device according to (1), further comprising means for processing a captured image and storing a result of identifying a leukocyte component into which leukocytes are finely classified. (3) The white blood cell fine classification device according to (2), wherein the means for identifying the finely classified white blood cell component has a learning identification function. (4) a step of staining leukocytes in a test solution, a step of inputting an image of the leukocytes, and sub-classifying the leukocytes using the image feature amount of the leukocyte image, and counting the number of the sub-classified leukocytes. A subclassification method for leukocytes, comprising a step of counting each leukocyte. (5) The white blood cell subclassification method according to (4), wherein the image feature amount is an optical feature amount. (6) The optical feature amount is an image obtained by performing color extraction that separates red, green, and blue into lightness and chromaticity, and performing binary image processing including filling in holes, writing line segments, and separating images. (5) the subclassification method of leukocytes, which is the feature quantity of (1). (7) The method of subclassifying leukocytes according to (4), wherein leukocyte components in urine are subdivided into dark cells, pale cells and glitter cells. (8) As a staining solution used in the step of staining leukocytes, Sternheimer-Malbin staining, Sternheimer staining,
Prescott-Brodie staining, Behre-Muhlberg staining, Sudan
III staining method, Lugol staining method, hemosiderin staining method, Papanic
olaou staining method, 4-chloro-1-naphthol method, Field staining method, Q
The subclassification method of leukocyte (7), which is a reagent containing one or more components used in at least one method selected from the group consisting of the uaglino-Flemans method, the Kaplow method, and the Sato-Sekiya method . (9) lymphocyte, monocyte, neutrophil,
The subclassification method of leukocytes according to (4), which is subclassified into eosinophils and basophils. (10) As a staining solution used in the step of staining leukocytes, Giemsa staining, Wright staining, May-Grunwald-Giemsa (Pappenheim) staining, Wright-Giemsa staining, Tomonaga method, Kaplow method, α-naphthyl. Butyrate staining method, α-naphthyl acetate staining method, naphthol chloroacetate staining method, diaminobenzidine
(DAB) staining method, 3-amino-9-ethylcarbazole (3AC) staining method, benzidine dihydrochloride (BDH) staining method, F
DA staining, α-naphthol staining, Fast garnet GBC staining, Fast red ITR staining, Pararosaniline staining, β
-Glucuronidase staining method, TdT staining method, Sudan black B
(SBB) Staining method, PAS reaction (Perriodic acid Shiff re
action) staining method, Simpson-Sabin staining method, brilliant cresyl blue staining method, Foilgen staining method and methyl green
The subclassification method of leukocytes according to (9), which is a reagent containing one or more components used in at least one method selected from the group consisting of pyronin staining. (11) The white blood cell fine classification apparatus according to any one of (1) to (3), which has a function of automatically focusing a sample image of white blood cells in a step of inputting an image of white blood cells.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A leukocyte fine classification apparatus according to the present invention is an imaging stage for imaging a test liquid on a light-transmitting plate or in a flow cell, and means for enlarging a sample image contained in the test liquid. Means for capturing the sample image, and means for processing the captured image to identify finely classified white blood cell components.
It is more preferable to have means for processing the captured image and storing the result of identifying the subdivided white blood cell component. Further, it is more preferable that the means for identifying the subdivided white blood cell component has a learning identification function.

In a preferred embodiment of the above-mentioned apparatus, the identification means has a learning recognition function.
Learning and identifying the range specification of the characteristic amount of leukocytes,
Having means for calculating the count of white blood cells based on an optical characteristic amount that is an image characteristic amount caused by white blood cells,
The magnification of the specimen image must be at least one, the material of the light-transmitting plate and the coated light-transmitting plate is glass or plastic, and the coated light-transmitting plate that covers the test liquid is formed integrally with the light-transmitting plate. It is characterized by having been performed.

Further, the subclassification method of leukocytes in the present invention comprises a step of staining leukocytes in a test solution, a step of inputting an image of the leukocytes, and a step of subdividing the leukocytes using the image feature amount of the leukocyte image. And a step of counting the number of the classified and sub-classified leukocytes for each sample. Here, the image feature amount refers to an index indicating information such as the shape and size of the obtained image, and specifically includes an optical feature amount. More specifically, a step of adding a reagent for assisting component identification to a test solution, an imaging stage for imaging the test solution on a light-transmitting plate or in a flow cell, and a sample image of white blood cells in the test solution The step of enlarging, the step of automatically focusing the sample image in the enlarged field of view, the step of capturing the sample image in the enlarged field of view, the step of processing the captured image to identify each classification, A step of storing the result of the classification, a step of repeating the above steps by changing the sample image until a preset number of images, a step of calculating an analysis result from all the identification results for the preset image,
And outputting the analysis result from an output device.

In the method of finely classifying leukocytes according to the present invention, after a test solution is placed on a light-transmitting plate, a sample image of white blood cells in the test solution is enlarged, and the enlarged sample image is automatically focused. Then, an enlarged specimen image is captured, and the captured image is processed and identified for each classification.

In the above method, a reagent for assisting the subclassification is added to the test solution, and the reagent for assisting the subclassification is obtained by the Sternheimer-Malbin staining method, St.
ernheimer staining, Prescott-Brodie staining, Behre-Muhl
berg staining, SudanIII staining, Lugol staining, hemosider
in staining method, Papanicolaou staining method, 4-chloro-1-naphthol
Method, Field staining method, Quaglino-Flemans method, Kaplow method, a reagent containing one or more components used in at least one of the Sato and Sekiya method,
Alternatively, Giemsa staining, Wright staining, May-Grunwald-Giemsa (Pappenheim) staining, Wright
Giemsa staining method, Tomonaga method, Kaplow method, α-naphthyl butyrate staining method, α-naphthyl acetate staining method,
Naphthol / chloroacetate staining method, diaminobenzidi
ne (DAB) staining method, 3-amino-9-ethylcarbazole (3AC)
Staining method, benzidine dihydrochloride (BDH) staining method,
FDA staining method, α-naphthol staining method, Fast garnet GBC
Staining method, Fast red ITR staining method, Pararosaniline staining method,
β-glucuronidase staining method, TdT staining method, Sudan black B (SBB) staining method, PAS reaction (Perriodic acid Shiff
reaction) A reagent containing one or more components used in at least one of staining method, Simpson-Sabin staining method, brilliant cresyl blue staining method, Foilgen staining method, and methyl green / pyronine staining method. There is,
The coated light-transmitting plate that covers the test liquid is formed integrally with the light-transmitting plate, and the light-transmitting plate is a slide glass,
More preferably, the material of the light-transmitting plate and / or the coated light-transmitting plate is glass or plastic or glass or plastic subjected to chemical treatment.

The leukocyte fine classification device of the present invention is characterized in that when leukocytes appear in a specimen, the leukocytes are converted into dark cells, pale cells and bright cells (gli cells).
It is a device that can be classified according to the type of staining and neutrophils, eosinophils, basophils, lymphocytes, and monocytes.

In the present invention, the sample to be analyzed is
Usually, urine (raw urine, concentrated urine, sediment after centrifugation) and blood, but other body fluids (serum, plasma, peritoneal fluid, semen, prostate fluid, synovial fluid, pleural effusion, ascites, secretions, etc.) It can be similarly applied to analysis of water and water quality inspection.

According to the present invention, since the object at the time of measurement is not a fluid, it can be focused, and since a light-transmitting plate is used without using a flow cell, a disposable one can be used. Since the possibility of carry-over of the sample and contamination by the staining solution is zero, it is possible to provide a highly precise and accurate measurement result.

[0021] The leukocyte fine classification apparatus of the present invention is characterized in that it has an imaging stage for imaging the test liquid on the light-transmitting plate, and the test liquid is placed on the light-transmitting plate. The material of the light-transmitting plate is not particularly limited as long as it has a light-transmitting property such as plastic (synthetic resin) or glass, but a slide glass is usually used. In the case of using a plastic, a plastic that has been subjected to a chemical treatment such as to improve hydrophilicity as necessary is used. Since the slide glass is single use (disposable),
The possibility of carryover of the sample and contamination by the stain is zero, and a highly reliable measurement result can be provided.

The material of the coated light-transmitting plate for coating the test liquid is not limited as in the case of the light-transmitting plate, but a cover glass is usually used. FIG. 1 is a perspective view of a cover glass-integrated slide glass 1 in which a coated light-transmitting plate is formed integrally with a light-transmitting plate. In FIG. 1A, a cover glass placed on a slide glass unit 2 is shown. Two opposing sides of the portion 3 are sealed with an adhesive 4 or the like, and the other two opposing sides are open. In FIG. 1B, three sides of a cover glass part 3 placed on the slide glass part 2 are sealed with an adhesive 4 or the like, and the other one is in an opened state. When the test liquid is dispensed from one of the opened sides, the test liquid is injected into the gap between the slide glass part 2 and the cover glass part 3 by capillary action. Accordingly, the slide glass 1 integrated with the cover glass can easily and accurately inject a predetermined amount, and can save labor for a complicated sample preparation step of setting the cover glass.

In addition, since the means for processing the captured image to finely classify the white blood cells has a learning recognition function, a highly accurate and precise measurement result is provided. The optical feature amount as the identification means is, for example, red,
Perform color extraction that separates green and blue into lightness and chromaticity, fill in the holes,
Identification is performed by learning the range designation of the feature amount of the image obtained by performing the binary image processing including the writing of the line segment and the separation of the image. here,
Image features include area, circularity coefficient, circle equivalent diameter, perimeter, absolute maximum length, Feret diameter X / Y ratio, maximum chord length X / Y
Ratio, short axis length / long axis length ratio, and the like.

The above-described image processing is based on the principle of classifying and analyzing the formed components in the test solution by the form using software. Therefore, the classification of leukocytes which may take various forms is not always sufficient. Thus, by adding a reagent such as a staining agent for assisting the subclassification to the test solution, the components similar to each other can be classified and the analytical performance can be improved.

The staining agent is not particularly limited, but Sternh is generally known for leukocytes in urine.
Eimer-Malbin staining method (SM staining method), Sternheimer staining method (S or NS staining or modified Sternheimer method), SudanIII staining method, Prescott-Brodie staining method, Behre
-Muhlberg staining (BM staining), Lugol staining, hemosideri
n staining method, Papanicolaou staining method, 4-chloro-1-naphthol
Method, Field staining method, Quaglino-Flemans method, Kaplow method, Satoh-Sekiya method, etc. (edited by Kiyoshi Okuda, “Clinical Inspection Atlas 1 Urine Sediment”, Medical and Dental Drug Publishing Co., Ltd .; “Clinical Inspection Technology Complete Book 3”
Blood tests; Izumihara Kanai, edited by Kanai, "Summary of Clinical Testing Methods," Kanehara Publishing Co., Ltd., monthly Medical Technology separate volume, all new staining methods, dyes and additives used in Medical and Dental Publishing Co., Ltd.) And a composition in which arbitrary components are combined from the dyes and additives constituting the above-described dyeing method.

On the other hand, for leukocytes in blood, Giemsa staining, Wright staining, May-Grunwald-Giemsa (Pappenheim) staining, Wright-Giemsa staining, Tomonaga method, Kaplow method, α-naphthyl butyrate staining Modulus, α
-Naphthyl acetate staining, naphthol chloroacetate staining, diaminobenzidine (DAB) staining,
3-amino-9-ethylcarbazole (3AC) staining method, benzidine
dihydrochloride (BDH) staining, FDA staining, α-naph
thol staining method, Fast garnet GBC staining method, Fast redITR staining method, Pararosaniline staining method, β-glucuronidase staining method, TdT staining method, Sudan black B (SBB) staining method, PAS
Reaction (Perriodic acid Shiff reaction) staining method, Simpson
-Sabin staining method, brilliant cresyl blue staining method, Foilgen staining method, methyl green / pyrronine staining, etc. (edited by Shiro Miwa, "Clinical Testing Techniques, Volume 3, Blood Test", Medical Shoin; Izumihara Kanai, edited by Kanai, " Summaries of Clinical Testing Methods ", Kinbara Publishing Co., Ltd., Monthly Medical Technology Supplement All new dyeing methods, including dyes and additives used in the Medical and Dental Publishing Co., Ltd. And a composition obtained by combining arbitrary components from additives.

In order to enhance the storage stability and preservative performance of the above-mentioned dye, generally known preservatives or antibacterial agents (various antibiotics, EDTA salts, boric acid, citric acid, NaN3, procrine, Benzisothiazolone, pyrithione, N-methylisothiazole, etc.), various buffers to keep the pH of the staining solution at an optimum pH, and various salts (EDT
A salt, phenate, oxalate, citrate, NaC
l, KCl, CaCl ₂ , AlCl ₃ etc.), various sugars (glucose, fructose, galactose, maltose, xylitol, sorbitol), cyclodextrins, glutaraldehyde, etc. to remove insoluble substances that hinder measurement. Various surfactants and enzymes may be added.

Further, the leukocyte fine classification apparatus of the present invention is preferably provided with a means for processing a captured image and storing the results of identification of various components, that is, an image storage device. Examples of the storage means include an auxiliary storage device such as a magneto-optical disk, a fixed disk, a digital video disk, and a CD-ROM having a large storage capacity. When a display is used as the output device, the menu selection allows the user to select the measurement result and image data as well as the time, the current status of the device (the measurement status of each sample, the scheduled end time of the measurement of each sample or the selected sample, or Function to display required remaining time, amount of waste liquid in waste liquid tank, remaining amount in pure water tank, remaining amount of each reagent, remaining amount of detergent, remaining number of slide glass), etc. A function of enlarging and displaying the image so that it can be read from the document may be added. Furthermore, "other" for leukocytes that cannot be classified by the identification means
It is also possible to add a function of classifying items into certain items, or a function of calling the image later and making a direct visual judgment by a technician, and adding or correcting the result of the judgment to data.

Also, a function may be added that allows remote control of selection of a display menu on the display and operation of the device using a remote control device. Further, when an accident occurs, when the waste liquid tank is full, when pure water, each reagent, detergent, slide glass, etc. are low, a warning is issued by a screen display, a sound or a signal. Functions may be added.

The subclassification method of leukocytes of the present invention can be easily performed by using the above-described apparatus. That is,
The step of automatically placing the test liquid on the light projecting plate is performed, the step of enlarging the specimen image of the white blood cells in the test liquid by the enlarging means is performed, and the white blood cell image is automatically focused by the imaging means, A step of capturing a sample image of a white blood cell is performed, and a step of processing the image captured by the identification unit to identify each type is performed. As described above, by using the leukocyte fine classification device of the present invention, the fine classification method of the present invention can be performed fully automatically.

[0031]

Embodiments of the apparatus and method of the present invention will be described below, but the present invention is not limited to the embodiments.

Embodiment 1 FIG. 2 is a configuration diagram of an embodiment of a leukocyte fine classification apparatus according to the present invention. A required amount is dispensed from the sample container containing the urine sample into the reaction tube 19 by a probe. At this time, sampling may be performed after the stirring operation in order to more uniformly collect the sample. When a sample container to which a barcode has been pasted is used, the barcode is read in advance by a barcode reading device. By using the barcode, you can read the presence / absence, type, and sample number of the sample,
It can be used for subsequent processing data and sample identification. Next, a required amount of the staining solution is dispensed into the reaction tube, and the leukocyte component is stained. After about 2 minutes, a required amount is taken out of the reaction tube 19 and dispensed onto a cover glass integrated slide glass 5 (hereinafter abbreviated as slide glass 5). The dyeing reaction time and reaction temperature may be set arbitrarily.

In FIG. 2, a slide glass 5 is placed on an XY table 6 which is an imaging stage, and the XY table 6 is driven so that the slide glass 5 can freely move and stop in the X coordinate axis direction and the Y coordinate direction. be able to. Light emitted from a lamp 7 as a light source travels on an optical axis, passes through a condenser lens 8, and is condensed on a test liquid on a slide glass 5. The formed image in the test liquid is enlarged by the objective lens 9 and adjusted to the image forming position by the automatic focusing mechanism. The image at this imaging position is projected as an image on the imaging surface of the CCD camera 10 and photoelectrically converted.

A driving circuit 11 for driving the XY table 6 is controlled by an image processing control circuit 12. The image processing control circuit 12 controls movement and stop of the XY table 6, and stops the XY table 6. At that time, CC
The D camera 10 is controlled to perform imaging. The image obtained by the CCD camera 10 is digitized by the A / D converter 13 and stored in the image memory 14. Image memory 1
After the storage in the field of view 4, the XY table 6 is moved to change the field of view, and the image is captured, A / D converted, and stored in the memory in the same process. repeat. Note that the movement of the XY table 6 for changing the field of view may be controlled to be performed after image analysis for one field of view is completed.

The writing to the image memory 14 and the control of the subsequent image processing are performed by the image processing control circuit 12, and the input of the image data stored in the image memory 14 to the feature extraction circuit 15 is controlled. . Feature extraction circuit 15
Is the characteristic amount of the image (the area formed, the circularity coefficient, the circle equivalent diameter, the perimeter, the absolute maximum length, the Feret diameter X / Y ratio, the maximum chord length X / Y ratio, the minor axis length / major axis Length ratio) as primary parameters. The image processing control circuit 12 inputs the primary parameters and the secondary parameters generated by a combination operation thereof to the identification operation circuit 16. The discrimination arithmetic circuit 16 classifies material components using a neural network. The neural network performs learning using a large amount of data based on expert judgment in advance, and the coupling coefficient between neurons is optimized. Therefore, a neural network operation is performed using the input feature parameters, and the automatic classification of the target tangible material can be performed. Further, instead of using a neural network, an automatic classification may be performed by using a statistical learning recognition method. If classification is not possible, a function may be added to classify the item into other component items, or call the image later and make direct judgment by the technician, and add or correct the judgment result to the data. .

The central control unit 17 stores the classification result and the image in the image storage device 18. The image storage device 18 generally uses a magneto-optical disk or fixed disk which has a large storage capacity and can be erased and written. The central control unit 17 integrates classification results for each visual field for all visual fields, and converts qualitative display items into qualitative data (-, ±, +, ++, +++, etc.) based on boundary values input in advance. After the conversion, the measurement result and the image are output to the output device 23 such as a display or a printer.

On the display, the menu displays the measurement result, image data, time, current device status (measurement status of each sample, scheduled end time or required remaining time of measurement of each sample or selected sample, waste liquid tank). Waste, pure tank remaining, reagents remaining, detergent remaining,
The function of displaying the remaining amount of the slide glass) or a function of enlarging the display so that only the selected and specified information can be read from a remote place may be added. In addition, a function may be added in which a display menu on the display or an operation of the device can be selected by remote control using a remote control device.

When an accident occurs, when the waste liquid tank is full, when pure water, reagents, detergents, or slide glass remains low, an alarm is generated by a screen display, a sound, or a signal. May be represented. In addition, an emergency test button is installed to respond to an urgent analysis, and when that button is pressed, the emergency sample is interrupted prior to the sample under analysis, or the analysis is paused. Alternatively, a function that can measure an urgent sample immediately may be added. In addition, all controls such as operation of each process, image processing, storage, calculation, and output can be performed by a program on a personal computer.

Example 2 [Urine Sample Collection Step] After stirring 200 samples of the undiluted urine without centrifugation, 0.75 ml of each sample is dispensed into a predetermined reaction tube. [Staining step] S (Sternheim)
er) Add 0.25 ml of staining solution and stir. [Specimen preparation step] 0.015 m from the solution containing the staining solution
1 is dispensed and dispensed into the interval portion of the cover glass-integrated slide glass to prepare a specimen. [Sample moving step] The prepared sample is set on the XY stage. The sample is moved to the imaging position by the XY stage. [Imaging Step] An enlarged image corresponding to a 10 × internal lens and a 400 × objective lens is viewed in 100 visual fields (strongly expanded visual field: HP) by a CCD camera set on an imaging stage.
F) Take an image. [Image processing / storage step] The captured image is stored in a magneto-optical disk or a fixed disk. The captured image is classified and counted into various components in urine using a learning recognition function, and the result is transmitted to a control calculation process. After storing the image, the XY stage moves to the next field of view, and proceeds to the focus step, the imaging step, and the storage step again, and repeats this until 100 fields are reached. [Control calculation step / output step] The result obtained from the image processing / storage step and the result obtained from the microbial mass analysis step are integrated and output to a display or a printer. Table 1
Shows an example of the output (printing), and the result in the high magnification field of view (HPF) is indicated by a symbol.

[0040]

[Table 1]

Table 2 shows the coincidence rate of the inspection result by the image processing in this embodiment. That is, each captured image was visually observed to determine whether or not the image coincided with the white blood cell fine classification classified and analyzed by the image processing. Table 2 summarizes the coincidence rates.

[0042]

[Table 2]

Example 3 [Specimen preparation process] EDTA-added blood 2 within 1 hour after blood collection
The sample is smeared on a clean slide glass to prepare a specimen. [Staining process] To the prepared specimen, 10 drops of May-Günwald solution (0.5% methylene blue eosinate in methyl alcohol solution) are added, and the mixture is allowed to stand for 1 to 2 minutes. Next, 10 drops of phosphate buffer were added, and the pre-solution was washed off with Giemsa staining solution.
Further, 3 ml of the same solution is applied and stained for 10 to 20 minutes. Finally, wash with running water. [Sample moving step] The prepared sample is set on the XY stage. The sample is moved to the imaging position by the XY stage. [Automatic focus step] The automatic focus adjustment mechanism focuses on the formed material. [Imaging Step] An enlarged image is captured by an objective lens using a CCD camera set on an imaging stage. [Image processing / storage step] The captured image is stored in a magneto-optical disk or a fixed disk. The captured image is classified and counted for each type of white blood cells using a learning recognition function, and the result is transmitted to a control calculation process. After the image is stored, the XY stage moves to the next field of view, and proceeds again to the focusing step, the imaging step, and the storage step.
Repeat until you have [Control calculation step / output step] The results for 200 pieces obtained from the image processing step are integrated and output to a display or a printer in the total number of white blood cells. Table 3 summarizes the coincidence rates of the inspection results by the image processing in the present embodiment. That is, each captured image was visually observed to determine whether or not the image coincided with the white blood cell fine classification classified and analyzed by the image processing. Table 3 summarizes the coincidence rates.

[0044]

[Table 3]

Example 4 [Urine Sample Collection Step] After stirring 200 samples of urine stock without centrifugation, 0.75 ml of each sample was dispensed into a predetermined reaction tube. [Staining step] S (Sternheim)
er) Add 0.25 ml of staining solution and stir. [Imaging step] The staining solution is dispensed into a light-transmitting plate or a flow cell, and an enlarged image corresponding to a 10 × internal lens and a 40 × objective lens is displayed in 100 visual fields (strong visual field: HPF) Take an image. [Image processing / storage step] The captured image is stored in a magneto-optical disk or a fixed disk. The captured image is classified and counted into various components in urine using a learning recognition function, and the result is transmitted to a control calculation process. After storing the image, the specimen image is moved, and the process again proceeds to the focusing step, the imaging step, and the storage step.

Table 4 shows the coincidence rates of the inspection results by the image processing in the present embodiment. That is, each captured image was visually observed to determine whether or not the image coincided with the white blood cell fine classification classified and analyzed by the image processing. Table 4 summarizes the coincidence rates.

[0047]

[Table 4]

[0048]

As described above, the use of the leukocyte fine classification apparatus of the present invention simplifies the work relating to the leukocyte fine classification in the test solution, reduces the burden on the laboratory technician, and improves the precision. , And can provide highly accurate analysis results.

[Brief description of the drawings]

FIG. 1 is a perspective view of a cover glass integrated slide glass 1. FIG.

FIG. 2 is a diagram showing an example of the leukocyte fine classification device of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Slide glass integrated with cover glass 2 Slide glass part 3 Cover glass part 4 Adhesive 5 Slide glass 6 XY table 7 Lamp 8 Condenser lens 9 Objective lens 10 CCD camera 11 Drive circuit 12 Image processing control circuit 13 A / D converter 14 Image memory 15 Feature extraction circuit 16 Identification operation circuit 17 Central control unit 18 Image storage device 19 Reaction tube 20 Luminometer 21 Integrator circuit 22 Operation circuit 23 Output device

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G01N 15/14 G01N 15/14 CD 33/48 33/48 M 33/483 33/483 C G06T 7 / 00 21/25 // G01N 21/25 (C12M 1/34 (C12M 1/34 C12R 1:91) C12R 1:91) (C12Q 1/00 Z (C12Q 1/00 C12R 1:91) C12R 1:91 (C12Q 1/06 (C12Q 1/06 C12R 1:91) C12R 1:91) G06F 15/62 395 (72) Inventor Keizo Yoneda 10-24 Toyo-cho, Tsuruga-shi, Fukui Toyobo Co., Ltd. Tsuruga Bio Inc. Inside the laboratory (72) Inventor Shinichi Tejima 2-2-2 Dojimahama, Kita-ku, Osaka-shi, Osaka Toyobo Co., Ltd. (72) Inventor Yoshiyuki Katsuma 2-2-2 Dojimahama, Kita-ku, Osaka-shi, Osaka Toyobo Corporation In-house F-term (reference) 2G045 AA03 AA25 BA14 BB24 CA11 CA25 CB03 FA11 FA16 GA03 GC22 JA01 JA03 2G059 AA06 BB14 CC16 DD03 DD13 EE13 FF01 GG10 HH02 JJ01 KK04 MM09 PP01 4B029 AA07 FA03 FA05 FA10 4B063 QA01 QA18 QQ08 QR58 QR66 QS36 QX01 5B057 AA10 BA02 CA01 CA06 CA12 CA16 DA13 DB02 DB06 DB08 DC03 DC04 DC25

Claims (11)

[Claims] An imaging stage for imaging a test solution on a light-transmitting plate or in a flow cell; a unit for enlarging a sample image contained in the test solution; a unit for imaging the sample image; Means for processing a captured image to identify a finely classified white blood cell component. 2. The white blood cell subclassifying apparatus according to claim 1, further comprising means for processing a captured image and storing a result obtained by identifying a white blood cell component obtained by subclassifying white blood cells. 3. The leukocyte fine classification device according to claim 2, wherein the means for identifying the finely classified white blood cell component has a learning identification function. 4. A step of staining leukocytes in a test solution, a step of inputting an image of the leukocytes, and sub-classifying the leukocytes using an image feature amount of the leukocyte image, and counting the number of the sub-classified leukocytes. A subclassification method for leukocytes, comprising the step of: 5. The method according to claim 4, wherein the image feature amount is an optical feature amount. 6. The optical feature amount is obtained by performing color extraction for separating red, green, and blue into lightness and chromaticity, filling in holes, writing line segments,
6. The white blood cell subclassification method according to claim 5, wherein the method is a feature amount of an image obtained by performing binary image processing including image separation. 7. A urine leukocyte component is converted to a dark cell (dark cel).
l), pale cells and glitter cells
The subclassification method of leukocytes according to claim 4, wherein the subclassification is performed. 8. The staining solution used in the step of staining leukocytes is Sternheimer-Malbin staining, Sternheimer staining.
Staining method, Prescott-Brodie staining method, Behre-Muhlberg staining method, SudanIII staining method, Lugol staining method, hemosiderin staining method, Papanicolaou staining method, 4-chloro-1-naphthol method, Fie
ld staining method, Quaglino-Flemans method, Kaplow method and Sato
The subclassification method of leukocytes according to claim 7, which is a reagent containing one or more components used in at least one method selected from the group consisting of the Sekiya method. 9. The method according to claim 9, wherein leukocyte components in blood are lymphocytes, monocytes,
The subclassification method of leukocytes according to claim 3, wherein the subclassification is performed into neutrophils, eosinophils, and basophils. 10. The staining solution used in the step of staining leukocytes includes Giemsa staining, Wright staining, May-Grunwald-Giemsa (Pappenheim) staining, Wright-Giemsa staining, Tomonaga method, Kaplow method, α- Naphthyl butyrate staining, α-naphthyl acetate staining, naphthol chloroacetate staining, diaminoben
zidine (DAB) staining method, 3-amino-9-ethylcarbazole (3
AC) staining method, benzidine dihydrochloride (BDH) staining method, FDA staining method, α-naphthol staining method, Fast garnet
GBC staining method, Fast red ITR staining method, Pararosaniline staining method, β-glucuronidase staining method, TdT staining method, Sudan black B (SBB) staining method, PAS reaction (Perriodic acid Sh
iff reaction) One or two types of components used in at least one method selected from the group consisting of a staining method, a Simpson-Sabin staining method, a brilliant cresyl blue staining method, a Feulgen staining method, and a methyl green / pyronine staining method The leukocyte fine classification device and the leukocyte fine classification method according to claim 9, which is a reagent containing the above. 11. The white blood cell subclassifying apparatus according to claim 1, further comprising a function of automatically adjusting the focus of a white blood cell sample image in the step of inputting a white blood cell image.