CN109255775A - A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically - Google Patents
A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically Download PDFInfo
- Publication number
- CN109255775A CN109255775A CN201810793315.4A CN201810793315A CN109255775A CN 109255775 A CN109255775 A CN 109255775A CN 201810793315 A CN201810793315 A CN 201810793315A CN 109255775 A CN109255775 A CN 109255775A
- Authority
- CN
- China
- Prior art keywords
- crypts
- image
- module
- microendoscopic
- optical fiber
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000013307 optical fiber Substances 0.000 title claims abstract description 40
- 230000002496 gastric effect Effects 0.000 title claims abstract description 31
- 238000004458 analytical method Methods 0.000 title claims abstract description 17
- 230000000877 morphologic effect Effects 0.000 claims abstract description 22
- 230000011218 segmentation Effects 0.000 claims abstract description 21
- 238000013139 quantization Methods 0.000 claims abstract description 16
- 238000001914 filtration Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 45
- 230000002708 enhancing effect Effects 0.000 claims description 33
- 238000004445 quantitative analysis Methods 0.000 claims description 18
- 239000000835 fiber Substances 0.000 claims description 11
- 238000006243 chemical reaction Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 241001584785 Anavitrinella pampinaria Species 0.000 claims description 6
- 238000005530 etching Methods 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 5
- 230000002093 peripheral effect Effects 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 230000002146 bilateral effect Effects 0.000 claims description 3
- 238000012887 quadratic function Methods 0.000 claims description 3
- 238000005260 corrosion Methods 0.000 claims 1
- 230000007797 corrosion Effects 0.000 claims 1
- 230000008030 elimination Effects 0.000 claims 1
- 238000003379 elimination reaction Methods 0.000 claims 1
- 210000000981 epithelium Anatomy 0.000 abstract description 9
- 230000003902 lesion Effects 0.000 abstract description 5
- 238000004195 computer-aided diagnosis Methods 0.000 abstract description 3
- 238000012549 training Methods 0.000 abstract description 3
- 230000006872 improvement Effects 0.000 description 9
- 238000003745 diagnosis Methods 0.000 description 5
- 230000003044 adaptive effect Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000000112 colonic effect Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 208000018522 Gastrointestinal disease Diseases 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 208000010643 digestive system disease Diseases 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 238000001839 endoscopy Methods 0.000 description 2
- 208000018685 gastrointestinal system disease Diseases 0.000 description 2
- 210000001035 gastrointestinal tract Anatomy 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 208000005718 Stomach Neoplasms Diseases 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 210000003238 esophagus Anatomy 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 206010017758 gastric cancer Diseases 0.000 description 1
- 210000004907 gland Anatomy 0.000 description 1
- 230000000762 glandular Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003703 image analysis method Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000003706 image smoothing Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 230000003211 malignant effect Effects 0.000 description 1
- 210000000214 mouth Anatomy 0.000 description 1
- 210000004877 mucosa Anatomy 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 201000011549 stomach cancer Diseases 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000008719 thickening Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- 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
- G06T2207/10061—Microscopic image from scanning electron microscope
-
- 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
- G06T2207/30092—Stomach; Gastric
-
- 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
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of gastrointestinal epithelial crypts structures based on optical fiber microendoscopic image to quantify analysis method and system automatically, and wherein system includes image filtering preprocessing module, contrast-enhancement module, crypts segmentation module, crypts morphological feature quantization modules.The present invention can apply in identifying benign, lesion epithelial tissue computer-aided diagnosis system, it can also handle immediately, real-time, the field diagnostic that scope doctor will be helped to realize high accuracy and consistency mitigate doctor's work and training burden, promote clinical efficiency.
Description
Technical field
The present invention relates to Medical Image Processing and application fields, specifically, being related to a kind of based on optical fiber microendoscopic figure
The gastrointestinal epithelial crypts structure of picture quantifies analysis method and system automatically.
Background technique
Gastroenteric tumor is malignant disease common in world wide, and China belongs to the High Risk For Gastric Cancer area, and early detection is
Improve the key of survival and quality of life.
Optical fiber microendoscopic based on fiber optic bundle is micro- interior with subcellular real time imagery ability, including confocal fluorescent
Mirror, high definition microendoscopic etc..In gastrointestinal mucosa screening application field, have many research shows that optical fiber microendoscopic will have
Clinical data can help the precancerous lesion and cancerous lesion of early detection GI epithelium.Therefore, optical fiber microendoscopic exists
There is huge clinical value in terms of gastrointestinal disease early detection.
As other advanced gastrointestinal tract endoscopic technics, the learning curve of optical fiber microendoscopic is long, the diagnosis effect of image
For fruit dependent on the pathological basis of doctor, high level training and rich experiences, these hinder its application in disorder in screening
It promotes.In recent years, it is considered being that one kind there are efficacious prescriptions using computer-aided diagnosis technology as the endoscopic image quantitative analysis of core
Method and attract attention.Early period, the quantitative analysis of optical fiber microendoscopic image concentrate on the diagnosis aspect in oral cavity, esophageal lesion, knot
Fruit shows image quantitative analysis as a kind of objective classification method, and it is aobvious to provide result auxiliary optical fiber accurate, that consistency is high
Micro- endoscopic image diagnosis, shows suitable with the evaluation capacity of veteran scope doctor.
In terms of gastrointestinal disease, glandular crypts are the important features of its diagnosis.Abnormal gastrointestinal epithelial tissue is disorder,
Crypts structure is elongated with epithelium irregular thickening, and the crypts of normal gastrointestinal epithelial tissue is similar round structure, and point
Cloth is more uniform, and therefore, compared with abnormal image, the crypts structure in normal picture is closer to circle, while area is less than normal,
Away from relatively uniform.The above-mentioned morphological feature general who has surrendered of quantization crypts helps distinguish between normal and abnormal gastrointestinal epithelial tissue.Crypts
Morphological feature has been confirmed to the value that gastrointestinal tract endoscopic image diagnoses in optical fiber microendoscopic equipment.Peieto et al. will
The quantization of crypts morphological feature is applied in the diagnosis of enteric epithelium optical fiber microendoscopic image, however the method for the research is to low
When the optical fiber microendoscopic image enhancement of contrast, need to select most preferably by multiple Self-adaptive strength contrast Enhancement test
Parameter combination increases the complexity of clinical application.Endoscopy application scenarios require image interpretation in real time, live, therefore
Prosthetic participates in, full automatic quantitative analysis has more advantage in clinical application.Currently, existing scientist is thin based on epithelium of esophagus
The morphological feature of born of the same parents and body of gland develops a set of Full-automatic optical fibre microendoscopic image quantization parser, however, entire method
The execution time need 52 seconds, be still difficult to meet requirement of real-time.
Therefore, it is necessary to develop a kind of more efficient convenient and fast image analysis method.
Summary of the invention
In order to solve the above-mentioned technical problem the present invention makes, a kind of more efficient, convenient and fast the purpose is to provide
Gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically.
To achieve the goals above, the present invention provides a kind of gastrointestinal epithelial crypts knots based on optical fiber microendoscopic image
Structure quantifies analysis method automatically, comprising the following steps:
(1) bilateral filtering method is used, pretreatment is filtered to original fiber microendoscopic image, to eliminate image slices
The problem of elementization, and keep the edge of image;
(2) image degree of comparing is enhanced:
(2.1) CLAHE method is used, enhancing for the first time is carried out to filtered image and is operated;
(2.2) nonlinear gray transformation is carried out using quadratic function, allows the histogram of image to spread toward both ends, widens image
The transformed gray scale of nonlinear gray is calculated as follows to each pixel on image in the contrast of foreground and background
Value obtains new images:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is processing
The original gray value of preceding pixel point;
(3) image is split, extracts crypts structure:
(3.1) enhanced image is switched to by binary image by the automatic division method based on threshold value, before being partitioned into
Scape (white, commonly using gray value 1 or 255 indicates) and background (black, commonly using gray value 0 indicates), wherein gray value is lower seemingly
Crypts part is divided into background, remaining is then prospect;
(3.3) gray value of foreground and background in image is replaced, at this point, prospect (white, commonly use gray value 1 or
255 indicate) represent crypts;
(3.4) by first corroding the opening operation expanded afterwards, in the case where being not obvious the area for changing foreground and background
The boundary in these regions is carried out smooth;
(3.5) edge of each complete crypts in image is extracted in binary image after singulation;
(4) according to following formula calculate crypts structure morphological feature, the area including each complete crypts structure, circularity and
Centroid distance;
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts
For the average value of all pixels point coordinate in its edge.
As a further improvement of the present invention, it is to keep histogram further equal that step (2.2) further includes step (2.3) later
Weighing apparatusization carries out CLAHE enhancing operation to image again.
As a further improvement of the present invention, during the CLAHE algorithm in step (2) is realized, using bilinear interpolation
To reduce the time complexity of algorithm, when interpolation, need to divide the image into that M column × N row is equal in magnitude, continuous nonoverlapping son
Image-region.
It as a further improvement of the present invention, further include step (3.2) between step (3.1) and (3.3) by first expanding
Cavity that may be present and narrow notch in foreground and background are filled in the closed operation of post-etching.
As a further improvement of the present invention, the M column × N row value 8 × 8.
As a further improvement of the present invention, the automatic division method based on threshold value in step (3.1) can use
Ostu method.
As a further improvement of the present invention, the crypts edge extracting in step (3.5) can be using based on connected domain
Contour tracing algorithm.
The present invention also provides a kind of automatic quantitative analysis system of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image
System, the system include image filtering preprocessing module, contrast-enhancement module, crypts segmentation module, the quantization of crypts morphological feature
Module;
Described image filter preprocessing module is for being filtered pretreatment to the original fiber microendoscopic image of loading;
The contrast-enhancement module increases for filtered image obtained to image filtering preprocessing module
Strong operation, improves the contrast of crypts and peripheral region;
The crypts segmentation module is split for enhanced image obtained to contrast-enhancement module, is extracted
Crypts structure out;
The crypts morphological feature quantization modules using crypts segmentation the obtained segmentation of module after binary image and
Crypts edge calculates the morphological feature of crypts structure, the area including each complete crypts structure, circularity and matter according to following formula
Heart distance etc.,
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts
For the average value of all pixels point coordinate in its edge.
As a further improvement of the present invention, the contrast-enhancement module includes two CLAHE enhancing modules and one
Nonlinear gray conversion module;
First CLAHE enhancing module uses CLAHE method, obtained to image filtering preprocessing module filtered
Image carries out enhancing operation for the first time, and enhancing result is supplied to nonlinear gray conversion module;
Nonlinear gray conversion module is calculated as follows each pixel on obtained enhanced image
The transformed gray value of nonlinear gray obtains new images, and is supplied to second CLAHE enhancing module:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is processing
The original gray value of preceding pixel point;
Second CLAHE enhancing module carries out CLAHE enhancing behaviour on the obtained transformed new images of nonlinear gray
Make, obtain final enhanced image, and is supplied to crypts segmentation module.
As a further improvement of the present invention, the crypts segmentation module includes image binaryzation module, and morphology closes fortune
Calculate module, gray inversion module, morphology opening operation module, crypts edge extracting module;
Described image binarization block is based on the automatic division method of threshold value by the obtained enhancing of contrast-enhancement module
Image afterwards switchs to binary image, be partitioned into prospect (white, commonly use gray value 1 or 255 indicate) and background (black, it is common
Gray value 0 indicates), wherein gray value is lower is divided into background like crypts part, remaining is then prospect;
The closing operation of mathematical morphology module is used in the obtained binary image of image binaryzation module, by first swollen
Cavity that may be present and narrow notch in foreground and background are filled in the closed operation of swollen post-etching;
The gray inversion module is by the ash of foreground and background in the obtained binary image of closing operation of mathematical morphology module
Angle value is replaced, at this point, prospect (white, common gray value 1 or 255 indicate) represents crypts;
The morphology opening operation module is used in the obtained binary image of gray inversion module, by first corroding
The opening operation expanded afterwards smoothly, disappear to the boundary in region in the case where being not obvious the area for changing foreground and background
The quantization of crypts morphological feature is supplied to except region burr that may be present or narrow connection, and by obtained binary image
Module;
The crypts edge extracting module utilizes binary image obtained by morphology opening operation module, extracts in image
The edge of each complete crypts, and it is supplied to crypts morphological feature quantization modules.
Optical fiber microendoscopic system uses probe of the fiber optic bundle as micro-imaging in the present invention, due to single in fiber optic bundle
Fiber core is different with covering light transmission rate, and acquired image has apparent pixelation (light and shade variation), reduces system
Resolution ratio.Commonly going pixelation method is gaussian filtering, although this method is simple, quick, effective, can be obscured simultaneously whole
A image and affect the resolution to details.Two-sided filter is the promotion version for Gaussian smoothing, not only allows for picture
Relationship of the element on space length, while joined the consideration of the gray scale similarity degree between pixel, therefore it is filtered to image smoothing
Have while wave and protects side characteristic.
In addition, the usual overall gray value of optical fiber microendoscopic image is low and contrast is unobvious, crypts is leveraged
It can identification.Therefore, the present invention needs to carry out enhancing operation to image before crypts segmentation of structures and quantization.It is existing to adopt
It is aobvious with self-adapting histogram equilibrium method (Adaptive Histogram Equalization, hereinafter referred to as AHE) enhancing optical fiber
Micro- endoscopic image contrast, but AHE has the problem of noise of same area in excessive enlarged drawing.In recent years, in low visibility
Art of image analysis, contrast limited adaptive histogram equalization (Contrast Limited Adaptive
HistogramEqualization, hereinafter referred to as CLAHE) it has received widespread attention, as the optimization of AHE, by counting
It (is usually taken before calculating Cumulative Distribution Function (Cumulative Distribution Function, CDF) with threshold value predetermined
Value is 3~4) histogram is cut, to achieve the purpose that limit noise amplification.
In field of medical image processing, threshold value is the common method in image segmentation, and many researchers propose automatic threshold
It is worth partitioning algorithm, the Otsu method that such as Otsu, N. are proposed is a kind of automatic threshold segmentation classic algorithm that image grayscale is adaptive.
When identifying the target in image, generally requires to make object edge tracking processing, be also Contour extraction, common method has connection
Domain contour following algorithm.
Compared with prior art, the present invention have it is following the utility model has the advantages that
The present invention provides a kind of automatic quantitative analysis side of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image
Method, the present invention can apply in identifying benign, lesion epithelial tissue computer-aided diagnosis system, additionally it is possible to place immediately
Reason, real-time, the field diagnostic that scope doctor will be helped to realize high accuracy and consistency mitigate doctor's work and training burden,
Promote clinical efficiency.
Detailed description of the invention
Fig. 1 is that the present invention is based on the streams that the gastrointestinal epithelial crypts structure of optical fiber microendoscopic image quantifies analysis method automatically
Cheng Tu;
Fig. 2 is the result schematic diagram that the present invention implements experimental image filter preprocessing;
Fig. 3 is to quantify analysis method automatically the present invention is based on the gastrointestinal epithelial crypts structure of optical fiber microendoscopic image to carry out
The flow chart of contrast enhancing;
Fig. 4 is to quantify analysis method automatically the present invention is based on the gastrointestinal epithelial crypts structure of optical fiber microendoscopic image to carry out
The flow chart of crypts segmentation;
Fig. 5 is that the present invention is based on the gastrointestinal epithelial crypts structures of optical fiber microendoscopic image to quantify analysis method image automatically
The result schematic diagram of quantitative analysis;
Fig. 6 is that the present invention is based on the knots that the gastrointestinal epithelial crypts structure of optical fiber microendoscopic image quantifies analysis system automatically
Structure schematic diagram.
Specific embodiment
It is next below with reference to the accompanying drawings that the present invention is further elaborated.
Join Fig. 1, present embodiments provide for a kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image is automatic
Quantitative analysis method, comprising the following steps:
(1) bilateral filtering method is used, pretreatment is filtered to original fiber microendoscopic image, to eliminate image slices
The problem of elementization, and keep the edge of image;
Filter result is as shown in Figure 2, wherein (a) is the original high definition microendoscopic image of people's normal colonic epithelia, (b)
(c) it is filtered image for the partial enlarged view of (a), (d) is the partial enlarged view of (c).
(2) image degree of comparing is enhanced:
(2.1) CLAHE method is used, enhancing for the first time is carried out to filtered image and is operated;
(2.2) nonlinear gray transformation is carried out using quadratic function, allows the histogram of image to spread toward both ends, widens image
The transformed gray scale of nonlinear gray is calculated as follows to each pixel on image in the contrast of foreground and background
Value obtains new images:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is processing
The original gray value of preceding pixel point;
(2.3) to equalize histogram further, CLAHE enhancing operation is carried out again to image;
Fig. 3 (e) is the transformed image of nonlinear gray, can be seen that the intensity profile of image still from its histogram (f)
Compared with concentration.After being enhanced using CLAHE, from result figure 3 (g)-(h) it can be seen that histogram further equalizes, concentrated from comparing
Gray scale interval become distribution in whole tonal ranges, the contrast of crypts and peripheral region is more obvious, it is possible thereby to
Find out analysis method of the invention compare the prior art method tool be significantly improved.
(3) join shown in Fig. 4, image be split, crypts structure is extracted:
(3.1) enhanced image is switched to by binary image by the automatic division method based on threshold value, before being partitioned into
Scape (white, commonly using gray value 1 or 255 indicates) and background (black, commonly using gray value 0 indicates), wherein gray value is lower seemingly
Crypts part is divided into background, remaining is then prospect;
(3.2) it by first expanding the closed operation of post-etching, fills that may be present empty and narrow in foreground and background
Notch;
(3.3) gray value of foreground and background in image is replaced, at this point, prospect (white, commonly use gray value 1 or
255 indicate) represent crypts;
(3.4) by first corroding the opening operation expanded afterwards, in the case where being not obvious the area for changing foreground and background
The boundary in these regions is carried out smooth;
(3.5) edge of each complete crypts in image is extracted in binary image after singulation;
(4) according to following formula calculate crypts structure morphological feature, the area including each complete crypts structure, circularity and
Centroid distance;
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts
For the average value of all pixels point coordinate in its edge.
Quantitative analysis result is as shown in Figure 5, wherein and (a) is the original high definition microendoscopic image of people's normal colonic epithelia,
(b) it is the centroid distance of each adjacent crypts, (c) is the area and circularities of each complete crypts.The crypts of normal colonic epithelia is answered
This is similar round structure, and distribution uniform is consistent with result (b)-(c).Present embodiment be by python language based on
It is executed on calculation machine (Intel (R) Xeon (R) E3-1230V2 3.30GHz CPU, 16GB RAM), from original fiber microendoscopic
Image graph 5 (a), which is input to, obtains Fig. 5 (b)-(c) result time-consuming~500ms, meets the real-time of endoscopy requirement, if using C+
The execution speed of+language, method will be promoted further, it can be seen that the analysis method in the present invention compared to it is more existing
There is significant progress in time-consuming.
Wherein, during the CLAHE algorithm in step (2) is realized, it is multiple to reduce the time of algorithm to use bilinear interpolation
Miscellaneous degree when interpolation, needs to divide the image into that M column × N row is equal in magnitude, continuous nonoverlapping sub-image area, and the M column ×
N row value 8 × 8.
The automatic division method based on threshold value in step (3.1) can use Ostu method.
Crypts edge extracting in step (3.5) can use the Contour tracing algorithm based on connected domain.
Join Fig. 6, it is automatic that present embodiment also provides a kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image
Quantitative analysis system, the system include image filtering preprocessing module, contrast-enhancement module, crypts segmentation module, crypts shape
State characteristic quantification module;
Described image filter preprocessing module is for being filtered pretreatment to the original fiber microendoscopic image of loading;
The contrast-enhancement module increases for filtered image obtained to image filtering preprocessing module
Strong operation, improves the contrast of crypts and peripheral region;
The crypts segmentation module is split for enhanced image obtained to contrast-enhancement module, is extracted
Crypts structure out;
The crypts morphological feature quantization modules using crypts segmentation the obtained segmentation of module after binary image and
Crypts edge calculates the morphological feature of crypts structure, the area including each complete crypts structure, circularity and matter according to following formula
Heart distance etc.,
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts
For the average value of all pixels point coordinate in its edge.
The contrast-enhancement module includes two CLAHE enhancing modules and a nonlinear gray conversion module;
First CLAHE enhancing module uses CLAHE method, obtained to image filtering preprocessing module filtered
Image carries out enhancing operation for the first time, and enhancing result is supplied to nonlinear gray conversion module;
Nonlinear gray conversion module is calculated as follows each pixel on obtained enhanced image
The transformed gray value of nonlinear gray obtains new images, and is supplied to second CLAHE enhancing module:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is processing
The original gray value of preceding pixel point;
Second CLAHE enhancing module carries out CLAHE enhancing behaviour on the obtained transformed new images of nonlinear gray
Make, obtain final enhanced image, and is supplied to crypts segmentation module.
The crypts segmentation module includes image binaryzation module, closing operation of mathematical morphology module, gray inversion module, form
Learn opening operation module, crypts edge extracting module;
Described image binarization block is based on the automatic division method of threshold value by the obtained enhancing of contrast-enhancement module
Image afterwards switchs to binary image, be partitioned into prospect (white, commonly use gray value 1 or 255 indicate) and background (black, it is common
Gray value 0 indicates), wherein gray value is lower is divided into background like crypts part, remaining is then prospect;
The closing operation of mathematical morphology module is used in the obtained binary image of image binaryzation module, by first swollen
Cavity that may be present and narrow notch in foreground and background are filled in the closed operation of swollen post-etching;
The gray inversion module is by the ash of foreground and background in the obtained binary image of closing operation of mathematical morphology module
Angle value is replaced, at this point, prospect (white, common gray value 1 or 255 indicate) represents crypts;
The morphology opening operation module is used in the obtained binary image of gray inversion module, by first corroding
The opening operation expanded afterwards smoothly, disappear to the boundary in region in the case where being not obvious the area for changing foreground and background
The quantization of crypts morphological feature is supplied to except region burr that may be present or narrow connection, and by obtained binary image
Module;
The crypts edge extracting module utilizes binary image obtained by morphology opening operation module, extracts in image
The edge of each complete crypts, and it is supplied to crypts morphological feature quantization modules.
It is described in an illustrative manner above with reference to attached drawing according to the present invention a kind of based on optical fiber microendoscopic figure
The gastrointestinal epithelial crypts structure of picture quantifies analysis method and system automatically.It will be understood by those skilled in the art, however, that for upper
State a kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image proposed by the invention quantify automatically analysis method and
System can also make various improvement on the basis of not departing from the content of present invention.Therefore, protection scope of the present invention should be by
The content of appended claims determines.
Claims (10)
1. a kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method automatically, it is characterised in that packet
Include following steps:
(1) bilateral filtering method is used, pretreatment is filtered to original fiber microendoscopic image, to eliminate image pixel
The problem of, and keep the edge of image;
(2) image degree of comparing is enhanced:
(2.1) CLAHE method is used, enhancing for the first time is carried out to filtered image and is operated;
(2.2) nonlinear gray transformation is carried out using quadratic function, allows the histogram of image to spread toward both ends, widens display foreground
With the contrast of background, the transformed gray value of nonlinear gray is calculated as follows to each pixel on image, is obtained
Obtain new images:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is picture before handling
The original gray value of vegetarian refreshments;
(3) image is split, extracts crypts structure:
(3.1) enhanced image is switched to by binary image by the automatic division method based on threshold value, it is (white is partitioned into prospect
Color, commonly using gray value 1 or 255 indicates) and background (black, commonly using gray value 0 indicates), wherein gray value is lower like crypts portion
Divide and be divided into background, remaining is then prospect;
(3.3) gray value of foreground and background in image is replaced, at this point, (white commonly uses 1 or 255 table of gray value to prospect
Show) represent crypts;
(3.4) by first corroding the opening operation expanded afterwards, in the case where being not obvious the area for changing foreground and background to this
The boundary in a little regions carries out smooth;
(3.5) edge of each complete crypts in image is extracted in binary image after singulation;
(4) morphological feature of crypts structure, the area including each complete crypts structure, circularity and mass center are calculated according to following formula
Distance;
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts is it
The average value of all pixels point coordinate in edge.
2. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 1 based on optical fiber microendoscopic image
Method, it is characterised in that: further include step (2.3) after step (2.2) be to equalize histogram further, to image again into
Row CLAHE enhancing operation.
3. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 1 based on optical fiber microendoscopic image
Method, it is characterised in that: during the CLAHE algorithm in step (2) is realized, it is multiple to reduce the time of algorithm to use bilinear interpolation
Miscellaneous degree when interpolation, needs to divide the image into that M column × N row is equal in magnitude, continuous nonoverlapping sub-image area.
4. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 1 based on optical fiber microendoscopic image
Method, it is characterised in that: further include the closed operation of step (3.2) by first expansion post-etching between step (3.1) and (3.3), fill out
Fill cavity that may be present and narrow notch in foreground and background.
5. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 2 based on optical fiber microendoscopic image
Method, it is characterised in that: the M column × N row value 8 × 8.
6. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 1 based on optical fiber microendoscopic image
Method, it is characterised in that: the automatic division method based on threshold value in step (3.1) can use Ostu method.
7. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 1 based on optical fiber microendoscopic image
Method, it is characterised in that: the crypts edge extracting in step (3.5) can use the Contour tracing algorithm based on connected domain.
8. a kind of gastrointestinal epithelial crypts structure as described in any one of claim 1-7 based on optical fiber microendoscopic image is certainly
Momentum analysis system, it is characterised in that: the system includes image filtering preprocessing module, contrast-enhancement module, crypts point
Cut module, crypts morphological feature quantization modules;
Described image filter preprocessing module is for being filtered pretreatment to the original fiber microendoscopic image of loading;
The contrast-enhancement module carries out enhancing behaviour for filtered image obtained to image filtering preprocessing module
Make, improves the contrast of crypts and peripheral region;
The crypts segmentation module is split for enhanced image obtained to contrast-enhancement module, is extracted hidden
Nest structure;
The crypts morphological feature quantization modules utilize the binary image and crypts after the crypts segmentation obtained segmentation of module
Edge, calculates the morphological feature of crypts structure according to following formula, the area including each complete crypts structure, circularity and mass center away from
From,
(a) area is defined as each intramarginal sum of all pixels of complete crypts;
(b) circularity is defined as:
C=4 π × area/perimeter square
Wherein, the sum of all pixels on each complete crypts edge Zhou Changwei;
(c) centroid distance is defined as the centroid distance of two neighboring complete crypts, wherein the center-of-mass coordinate of each complete crypts is it
The average value of all pixels point coordinate in edge.
9. the gastrointestinal epithelial crypts structure automatic quantitative analysis side according to claim 8 based on optical fiber microendoscopic image
The system of method, it is characterised in that: the contrast-enhancement module includes that two CLAHE enhancing modules and a nonlinear gray become
Change the mold block;
First CLAHE enhancing module uses CLAHE method, filtered image obtained to image filtering preprocessing module
Enhancing operation for the first time is carried out, and enhancing result is supplied to nonlinear gray conversion module;
Nonlinear gray conversion module is calculated as follows each pixel non-thread on obtained enhanced image
Property greyscale transformation after gray value, obtain new images, and be supplied to second CLAHE enhancing module:
Wherein, coordinate (x, y) indicates the coordinate of pixel,For the new gray value of pixel, f (x, y) is processing
The original gray value of preceding pixel point;
Second CLAHE enhancing module carries out CLAHE enhancing operation on the obtained transformed new images of nonlinear gray, obtains
To final enhanced image, and it is supplied to crypts segmentation module.
10. the gastrointestinal epithelial crypts structure automatic quantitative analysis according to claim 8 based on optical fiber microendoscopic image
The system of method, it is characterised in that: the crypts segmentation module includes image binaryzation module, closing operation of mathematical morphology module, ash
Spend reversal block, morphology opening operation module, crypts edge extracting module;
Described image binarization block is obtained enhanced by contrast-enhancement module based on the automatic division method of threshold value
Image switchs to binary image, is partitioned into prospect (white, commonly using gray value 1 or 255 indicates) and background (black, common gray scale
Value 0 indicates), wherein gray value is lower is divided into background like crypts part, remaining is then prospect;
The closing operation of mathematical morphology module is used in the obtained binary image of image binaryzation module, after first expanding
Cavity that may be present and narrow notch in foreground and background are filled in the closed operation of corrosion;
The gray inversion module is by the gray value of foreground and background in the obtained binary image of closing operation of mathematical morphology module
It is replaced, at this point, prospect (white, common gray value 1 or 255 indicate) represents crypts;
The morphology opening operation module is used in the obtained binary image of gray inversion module, by swollen after first corroding
Swollen opening operation carries out smooth, elimination area in the case where being not obvious the area for changing foreground and background to the boundary in region
Domain burr that may be present or narrow connection, and obtained binary image is supplied to crypts morphological feature quantization mould
Block;
The crypts edge extracting module utilizes binary image obtained by morphology opening operation module, extracts each complete in image
The edge of whole crypts, and it is supplied to crypts morphological feature quantization modules.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810793315.4A CN109255775A (en) | 2018-07-18 | 2018-07-18 | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810793315.4A CN109255775A (en) | 2018-07-18 | 2018-07-18 | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109255775A true CN109255775A (en) | 2019-01-22 |
Family
ID=65048926
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810793315.4A Pending CN109255775A (en) | 2018-07-18 | 2018-07-18 | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109255775A (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110006898A (en) * | 2019-03-28 | 2019-07-12 | 上海工程技术大学 | The blended ratio measuring method of polyester fiber based on circularity algorithm |
CN110136161A (en) * | 2019-05-31 | 2019-08-16 | 苏州精观医疗科技有限公司 | Image characteristics extraction analysis method, system and device |
CN110264418A (en) * | 2019-05-31 | 2019-09-20 | 苏州精观医疗科技有限公司 | Method for enhancing picture contrast, system and device |
CN111563517A (en) * | 2020-04-20 | 2020-08-21 | 腾讯科技(深圳)有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
CN112089397A (en) * | 2020-07-05 | 2020-12-18 | 唐婧 | Inflammation analysis system using swallowable detection platform |
CN113177949A (en) * | 2021-04-16 | 2021-07-27 | 中南大学 | Large-size rock particle feature identification method and device |
CN114022879A (en) * | 2020-07-15 | 2022-02-08 | 苏州精观医疗科技有限公司 | Squamous cell structure enhancement method based on optical fiber endomicroscopy image |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
CN107180421A (en) * | 2016-03-09 | 2017-09-19 | 中兴通讯股份有限公司 | A kind of eye fundus image lesion detection method and device |
-
2018
- 2018-07-18 CN CN201810793315.4A patent/CN109255775A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
CN107180421A (en) * | 2016-03-09 | 2017-09-19 | 中兴通讯股份有限公司 | A kind of eye fundus image lesion detection method and device |
Non-Patent Citations (3)
Title |
---|
夏敏捷 等: "《Python程序设计应用教程》", 北京:中国铁道出版社, pages: 286 - 288 * |
张萌等: "《基于生猪外形特征图像的瘦肉率估测方法》", 《农业工程学报》, vol. 33, no. 12, pages 308 - 314 * |
李洪生: "《多光子显微成像技术用于结直肠腺癌发展进程监测的研究》", 《中国优秀硕士学位论文全文数据库》, no. 05, pages 30 - 35 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110006898A (en) * | 2019-03-28 | 2019-07-12 | 上海工程技术大学 | The blended ratio measuring method of polyester fiber based on circularity algorithm |
CN110136161A (en) * | 2019-05-31 | 2019-08-16 | 苏州精观医疗科技有限公司 | Image characteristics extraction analysis method, system and device |
CN110264418A (en) * | 2019-05-31 | 2019-09-20 | 苏州精观医疗科技有限公司 | Method for enhancing picture contrast, system and device |
CN111563517A (en) * | 2020-04-20 | 2020-08-21 | 腾讯科技(深圳)有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
CN111563517B (en) * | 2020-04-20 | 2023-07-04 | 腾讯科技(深圳)有限公司 | Image processing method, device, electronic equipment and storage medium |
CN112089397A (en) * | 2020-07-05 | 2020-12-18 | 唐婧 | Inflammation analysis system using swallowable detection platform |
CN112089397B (en) * | 2020-07-05 | 2021-07-13 | 陈伟 | Inflammation analysis system using swallowable detection platform |
CN114022879A (en) * | 2020-07-15 | 2022-02-08 | 苏州精观医疗科技有限公司 | Squamous cell structure enhancement method based on optical fiber endomicroscopy image |
CN114022879B (en) * | 2020-07-15 | 2024-08-23 | 苏州欧谱曼迪科技有限公司 | Squamous cell structure enhancement method based on optical fiber microscopy endoscope image |
CN113177949A (en) * | 2021-04-16 | 2021-07-27 | 中南大学 | Large-size rock particle feature identification method and device |
CN113177949B (en) * | 2021-04-16 | 2023-09-01 | 中南大学 | Large-size rock particle feature recognition method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109255775A (en) | A kind of gastrointestinal epithelial crypts structure based on optical fiber microendoscopic image quantifies analysis method and system automatically | |
CN111899229A (en) | Advanced gastric cancer auxiliary diagnosis method based on deep learning multi-model fusion technology | |
CN110276356A (en) | Eye fundus image aneurysms recognition methods based on R-CNN | |
TWI696145B (en) | Colonoscopy image computer-aided recognition system and method | |
CN108510502A (en) | Melanoma picture tissue segmentation methods based on deep neural network and system | |
CN110490892A (en) | A kind of Thyroid ultrasound image tubercle automatic positioning recognition methods based on USFaster R-CNN | |
CN109635871B (en) | Capsule endoscope image classification method based on multi-feature fusion | |
CN110136161A (en) | Image characteristics extraction analysis method, system and device | |
CN106683080A (en) | Retinal fundus image preprocessing method | |
CN106780439B (en) | A method of screening eye fundus image | |
CN105657580A (en) | Capsule endoscopy video summary generation method | |
CN109087310A (en) | Dividing method, system, storage medium and the intelligent terminal of Meibomian gland texture region | |
CN104318565A (en) | Interactive method for retinal vessel segmentation based on bidirectional region growing of constant-gradient distance | |
CN114332910A (en) | Human body part segmentation method for similar feature calculation of far infrared image | |
Bernal et al. | A region segmentation method for colonoscopy images using a model of polyp appearance | |
CN113139929A (en) | Gastrointestinal tract endoscope image preprocessing method comprising information screening and fusion repairing | |
CN117593540A (en) | Pressure injury staged identification method based on intelligent image identification technology | |
CN117593317A (en) | Retina blood vessel image segmentation method based on multi-scale dilation convolution residual error network | |
CN111292285B (en) | Automatic screening method for diabetes mellitus based on naive Bayes and support vector machine | |
CN111899272A (en) | Fundus image blood vessel segmentation method based on coupling neural network and line connector | |
CN112837259A (en) | Image processing method for skin pigment pathological change treatment effect based on feature segmentation | |
CN116434920A (en) | Gastrointestinal epithelial metaplasia progression risk prediction method and device | |
CN111062909A (en) | Method and equipment for judging benign and malignant breast tumor | |
Xue et al. | Automatic extraction of mosaic patterns in uterine cervix images | |
Resita et al. | Color RGB and structure GLCM method to feature extraction system in endoscope image for the diagnosis support of otitis media disease |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 215127 No. 104-106, Wanlong building, No. 29, Xinfa Road, Suzhou area, China (Jiangsu) pilot Free Trade Zone, Suzhou City, Jiangsu Province, 1st floor, auxiliary building (this address shall not be engaged in retail) Applicant after: Suzhou Oupu Mandi Technology Co.,Ltd. Address before: 215028 unit 302, building A6, biomedical industrial park, 218 Xinghu street, Suzhou Industrial Park, Jiangsu Province Applicant before: SUZHOU JINGGUAN MEDICAL TECHNOLOGY CO.,LTD. |