CN112801939A - Method for improving index accuracy of pathological image KI67 - Google Patents

Method for improving index accuracy of pathological image KI67 Download PDF

Info

Publication number
CN112801939A
CN112801939A CN202011615824.1A CN202011615824A CN112801939A CN 112801939 A CN112801939 A CN 112801939A CN 202011615824 A CN202011615824 A CN 202011615824A CN 112801939 A CN112801939 A CN 112801939A
Authority
CN
China
Prior art keywords
positive tumor
lymphocyte
lymphocytes
tumor cell
tumor cells
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.)
Granted
Application number
CN202011615824.1A
Other languages
Chinese (zh)
Other versions
CN112801939B (en
Inventor
蔡佳桐
杨林
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Diyingjia Technology Co ltd
Original Assignee
Hangzhou Diyingjia Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hangzhou Diyingjia Technology Co ltd filed Critical Hangzhou Diyingjia Technology Co ltd
Priority to CN202011615824.1A priority Critical patent/CN112801939B/en
Publication of CN112801939A publication Critical patent/CN112801939A/en
Application granted granted Critical
Publication of CN112801939B publication Critical patent/CN112801939B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
    • 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
    • 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
    • G06T2207/30242—Counting objects in image

Landscapes

  • Engineering & 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)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

The invention relates to a method for improving the accuracy of a pathological image KI67 index, which comprises the steps of identifying lymphocytes and positive tumor cells on a pathological section under the current visual field through a computer-aided discrimination algorithm, correcting the result of the computer-aided interpretation algorithm by adopting a neighborhood voting method through the correlation between the cell distance and the pigment intensity of the lymphocytes and the positive tumor cells, voting by using the positive tumor cells as the lymphocytes, wherein the voting principle is as follows: if the lymphocyte is close to most of the positive tumor cells in distance and color, the lymphocyte has a high probability of being a positive tumor cell, so that the part of the tumor cells which are misjudged as the lymphocyte by a computer-aided algorithm are corrected, a more accurate cell counting result is obtained, and the accuracy of KI67 interpretation is improved.

Description

Method for improving index accuracy of pathological image KI67
Technical Field
The invention relates to the technical field of medicine, in particular to a method for improving the accuracy of an index of a pathological image KI 67.
Background
With the penetration of digital construction in the medical industry, the application of computer-aided diagnosis and interpretation in clinic is increasing. Computer-aided diagnosis has a very high effect in some clinical application scenes with intensive and repeated labor, large film reading amount and the like. Ki67 pathological section interpretation is one of these scenes. KI67 is a pathological immunohistochemical technique, and clinically KI67 index is used to evaluate the malignancy of tumors. According to the standard, a pathologist needs to count the negative tumors and the positive tumors in not less than ten under-lens visual fields, then calculate the ratio of the positive tumors to all tumor cells and average the ratio to serve as a KI67 index. However, due to the similarity of the morphology between lymphocytes and positive tumor cells in pathological images, the algorithm that is easy to cause computer-aided interpretation often misjudges two types of cells, so that the calculation of the KI67 index has serious deviation.
Disclosure of Invention
The present application is proposed to solve the above-mentioned technical problems. The application provides a method and a computer readable medium for improving the accuracy of the index of a pathological image KI67, wherein a correction part is misjudged as tumor cells of lymphocytes by a computer-aided algorithm, so that a more accurate cell counting result is obtained, and the accuracy of KI67 interpretation is improved.
According to one aspect of the application, a method for improving the accuracy of an index of a pathology image KI67 is provided, which includes:
s10, recognizing lymphocytes and positive tumor cells on the pathological section under the current visual field through a computer-aided discrimination algorithm, wherein the total number of the lymphocytes is n, and the total number of the positive tumor cells is m;
s20, screening central points of all the lymphocytes and positive tumor cells;
s30, extracting information of each lymphocyte and positive tumor cell, wherein the information comprises the abscissa and the ordinate of the cell and the average value of the red channel pixel intensity on a circular pixel area taking the central point as the center;
s40, calculating the pixel distance between the ith lymphocyte and the jth positive tumor cell according to the information
Figure BDA0002874890960000011
Calculating the pixel intensity distance of the red channel between the ith lymphocyte and the jth positive tumor cell according to the information
Figure BDA0002874890960000021
Wherein i is more than or equal to 1 and less than or equal to n, and j is more than or equal to 1 and less than or equal to m;
s50, according to the pixel distance
Figure BDA0002874890960000022
And pixel intensity distance
Figure BDA0002874890960000023
Calculating the combined distance between the ith lymphocyte and the jth positive tumor cell
Figure BDA0002874890960000024
Generating a combinatorial matrix Mcombine，McombineThe ith row and the jth column of (A) elements are
Figure BDA0002874890960000025
S60, combining the matrix McombineAdding the values of the second dimension to obtain an n-dimensional vector VvoteThe result of voting on n lymphocytes by all positive tumor cells in the current visual field image is obtained;
s70, judging whether the voting result is effective, if so, correcting the lymphocyte misjudged as the positive tumor cell into the positive tumor cell;
and S80, calculating a KI67 index according to the corrected number of the positive tumor cells.
Further, the S70 specifically includes:
judging whether the discrimination result of the computer-aided discrimination contains a category confidence level,
if yes, obtaining the confidence coefficient of the categories of the m lymphocytes, and marking as VptJudging the n lymphocytes one by one, and for the ith lymphocyte, if so
Figure BDA0002874890960000026
The voting result is considered to be valid, and the i-th lymphocyte is a positive tumor cell misjudged as a lymphocyte, so that the classification of the i-th lymphocyte is corrected to be a positive tumor cell, and if so, the classification is judged to be valid
Figure BDA0002874890960000027
The voting result is considered to be invalid;
if not, setting a standard confidence coefficient Vt，VtAs a constant, judging n lymphocytes one by one, and for the ith lymphocyte, if
Figure BDA0002874890960000028
The voting result is considered to be valid, and the i-th lymphocyte is a positive tumor cell misjudged as a lymphocyte, so that the classification of the i-th lymphocyte is corrected to be a positive tumor cell, and if so, the classification is judged to be valid
Figure BDA0002874890960000029
The voting result is considered invalid.
Further, the
Figure BDA00028748909600000210
Figure BDA00028748909600000211
Expressed as the combined distance between the ith lymphocyte and the jth positive tumor cell, where α, β, dstd1And is dstd2Are all constants, alpha is used for adjusting the value of the result,
β is used to adjust the relative weights of the pixel distance and pixel intensity distance.
Further, the computer-aided discrimination algorithm includes a morphology-based conventional image algorithm and a deep learning algorithm.
Further, the lymphocyte and the positive tumor cell are both cell types output by the computer-aided identification algorithm.
According to yet another aspect of the application, a computer-readable medium is provided, on which computer program instructions are stored, which, when executed by a processor, cause the processor to perform the method for improving the accuracy of the KI67 index of a pathology image.
Compared with the prior art, the method for improving the accuracy of the KI67 index of the pathological image and the electronic device/computer readable medium can correct the result of the computer-aided interpretation algorithm by using the neighborhood voting method through the correlation between the cell distance and the pigment intensity of the lymphocyte and the positive tumor cell, so that a more accurate cell counting result is obtained, and the accuracy of KI67 interpretation is improved.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.
FIG. 1 is a flow chart of a method of an embodiment of the present application;
FIG. 2 shows the results of cell localization and classification of image under-mirror field image of computer-aided interpretation KI 67;
FIG. 3 is a view under a pathological section mirror of KI 67;
FIG. 4 is a view of the visual results visualized under the mirror using computer-aided interpretation KI67 for FIG. 3;
FIG. 5 is a visualization of the results after modification using the present application-the original algorithm does not contain a category confidence;
FIG. 6 is a visualization of the results after modification using the present application-the original algorithm contains class confidence;
FIG. 7 is a diagram of a small result of the truncation of FIG. 4;
FIG. 8 is a graph of the results of FIG. 7 modified using the present application-the original algorithm contains class confidence;
FIG. 9 is a diagram of the results of FIG. 7 modified using the present application-the original algorithm does not contain a class confidence.
Detailed Description
Hereinafter, example embodiments of the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
Summary of the application
KI67 is a pathological immunohistochemical technique, and clinically KI67 index is used to evaluate the malignancy of tumors. According to the criterion, a pathologist needs to count the negative tumor cells and the positive tumor cells in not less than ten under-lens visual fields, then calculate the ratio of the positive tumor cells to all the tumor cells, and average the ratio to be used as a KI67 index. At present, the traditional algorithm and the deep learning algorithm based on morphology are mainly used in the algorithm of computer-aided interpretation. Both of them can locate all cells and distinguish cell types in the under-mirror field, as shown in FIG. 2, where blue dots are negative fibers; green dots are negative tumor cells; the brown spots are positive fibers, the yellow spots are lymphocytes, the red spots are positive tumor cells, and the purple spots are other cells. However, due to the morphological similarity between lymphocytes and positive tumor cells in pathological images, the algorithms for computer-aided interpretation tend to misjudge both types of cells. The KI67 index is considered as an indicator of the proliferation of tumor cells in many neoplastic pathologies, with higher KI67 indices indicating faster tumor cell proliferation. The KI67 index is generally used for tumor grading, to assess patient prognosis and to determine whether a patient is susceptible to chemotherapy, among other things. Misjudging lymphocytes as positive tumor cells can affect the accuracy of the KI67 index, so that the judgment of tumor diseases is biased.
Aiming at the technical problem, the conception of the application is as follows: empirically, in more prominent cancer groove areas, the number of lymph is lower. However, in the visual result visualization under the computer-aided interpretation KI67 mirror, the visual result visualization often appears in the obvious cancer groove area, and the lymphocyte points account for a large proportion, which is an obvious error and can not be accepted by doctors. In fact, the cancer cell area generally has only a few lymphocytes, and most of the lymphocytes are misdiagnosed as positive tumor cells, and besides, the positive tumor cells read by computer-aided judgment are basically accurate. The application thus votes for lymphocytes using positive tumor cells, the voting principle being: if the lymphocyte is close to most positive tumor cells and the color of the lymphocyte is similar, the lymphocyte has a high probability of being a positive tumor cell. Two classes of cells, both computer-aided interpretation classes, non-authentic classes, are mentioned below. For positive tumor cells, the dominant hue is red, so the red channel is selected, and the pixel intensity pair of the red channel is used to determine whether the cell type is misjudged.
Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.
Exemplary method
A method for improving the accuracy of an index of a pathology image KI67, as shown in fig. 1, includes the following steps:
s10, recognizing lymphocytes and positive tumor cells on the pathological section under the current visual field through a computer-aided discrimination algorithm, wherein the total number of the lymphocytes is n, and the total number of the positive tumor cells is m;
the computer-aided identification algorithm can be a traditional image algorithm based on morphology, and can also be a deep learning algorithm. The present invention does not require the recognition result obtained by which manner. But the results contained at least two categories, positive tumor, lymphocytes. The subsequent correction steps are developed around these two types of cells, both of which, when referred to below, are computer-aided interpretation classes, non-authentic.
S20, screening central points of the lymphocytes and the positive tumor cells;
s30, extracting information of each lymphocyte and positive tumor cell, wherein the information comprises the abscissa and the ordinate of the cell and the average value of the red channel pixel intensity on a circular pixel area taking the central point as the center;
this step stores information for each lymphocyte and positive tumor cell for use in subsequent steps, including the location of the cell's center point (abscissa and ordinate), and the red channel over a region centered on the cell's center point and having a radius of r (r generally does not exceed the average cell radius)Is calculated as the average of the pixel values of (1). The horizontal and vertical coordinates of the cells are used to calculate the distance between two cells, and the mean value of the pixel intensity of the red channel is used to measure the difference in color between two cells (for positive tumor cells, the dominant hue is red, so the red channel is selected). S40, calculating the pixel distance between the ith lymphocyte and the jth positive tumor cell according to the information
Figure BDA0002874890960000051
Generating a matrix
Figure BDA0002874890960000052
Calculating the pixel intensity distance of the red channel between the ith lymphocyte and the jth positive tumor cell according to the information
Figure BDA0002874890960000053
Generating a matrix
Figure BDA0002874890960000054
Wherein i is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m,
Figure BDA0002874890960000055
and
Figure BDA0002874890960000056
all are n x m dimensional matrices;
Figure BDA0002874890960000057
Figure BDA0002874890960000061
s50, according to the pixel distance
Figure BDA0002874890960000062
And pixel intensity distance
Figure BDA0002874890960000063
Calculating the combined distance between the ith lymphocyte and the jth positive tumor cell to generate a combined matrix Mcombine；
Figure BDA0002874890960000064
McombineRow i and column j of
Figure BDA0002874890960000065
Expressed as the combined distance between the ith lymphocyte and the jth positive tumor cell, wherein alpha, beta, and d are in the above formulastd1And is dstd2Are constants, alpha is used to adjust the magnitude of the resulting value, and beta is used to adjust the relative weights of the pixel distance and the pixel intensity distance. The above values are determined by the performance of the algorithm in the experiment, dstd1And dstd2The presence of (a) allows the neighborhood voting method to assign higher voting weights to positive tumor cells that are close to lymphocytes.
S60, combining the matrix McombineSumming the values of the second dimension to obtain an n-dimensional vector, namely a voting result of all positive tumor cells in the current visual field image on n lymphocytes, taking sigmoid as the result,
Figure BDA0002874890960000066
is marked as Vvote,VvoteAlso an n-dimensional vector.
S70, judging whether the discrimination result of the computer-aided discrimination contains a category confidence (namely the probability that the algorithm recognizes the cell as the category) or not,
if yes, obtaining the confidence coefficient of the category of the n lymphocytes, and marking as VptJudging the n lymphocytes one by one, and for the ith lymphocyte, if so
Figure BDA0002874890960000067
The voting result is considered to be valid, and the ith lymphocyte is a positive tumor cell misjudged as a lymphocyte becauseAnd the classification of the ith lymphocyte is corrected to be a positive tumor cell if
Figure BDA0002874890960000068
The voting result is considered to be invalid;
if not, setting a standard confidence coefficient Vt，VtAs a constant, judging n lymphocytes one by one, and for the ith lymphocyte, if
Figure BDA0002874890960000071
The voting result is considered to be valid, and the i-th lymphocyte is a positive tumor cell misjudged as a lymphocyte, so that the classification of the i-th lymphocyte is corrected to be a positive tumor cell, and if so, the classification is judged to be valid
Figure BDA0002874890960000072
The voting result is considered to be invalid; i.e. the lymphocyte is not misjudged, the classification of the resulting spot will not be modified.
And S80, calculating a KI67 index according to the corrected number of the positive tumor cells.
As shown in fig. 3, which is a 1080 × 1920-sized image of the under-lens view of the KI67 pathological section, the large brown area in the upper left region of the image is the cancer groove area. FIG. 4 is a visualization of the results of the computer-aided interpretation KI67 view under the mirror. As can be seen from fig. 4, the computer-aided interpretation algorithm identifies more than half of the brown cells (true positive tumor cells) as lymphocytes (yellow dots), which are not present in the actual pathological section interpretation, and thus there are cases where the cells are misjudged.
If the computer-aided interpretation result does not include the category confidence, after correction by the method disclosed in the present application, almost all the yellow spots in the cancer cell area are corrected to red spots, i.e., all the lymphocytes in the cancer cell area are corrected to positive tumor cells (as shown in fig. 5).
If the computer-aided interpretation result includes a category confidence, most of the yellow spots are corrected to red spots, i.e., most of the lymphocytes are corrected to positive tumor cells (e.g., positive tumor cells) after correction by the method disclosed in the present applicationAs shown in fig. 6). Unlike the previous calibration method, some lymphocyte spots still remain in the patch. These lymph nodes are retained because of the higher probability of being considered lymphocytes by computer-aided judgment, i.e.
Figure BDA0002874890960000073
Figure BDA0002874890960000074
The voting result is invalid. After being confirmed by a pathologist, the cells are smaller than peripheral cells, are irregular polygons in shape and accord with characteristics of lymphocytes, and the algorithm is reasonable in correcting the lymphocytes.
The voting revision process is described below:
as shown in figure 7, V obtained after sigmoid is taken from the voting results of all positive tumor cells on lymphocyte spots 1-4 in the current visual field imagevote＝[0.9,0.81,0.83,0.58]The discrimination result of the computer-aided discrimination includes a category confidence Vpt＝[0.8,0.9,0.7,0.95]In comparison, the voting results of the lymphocyte spots 1 and 3 are greater than the corresponding category confidence degrees, and the voting results are considered to be valid, the lymphocyte spots 1 and 3 are positive tumor cells misjudged as lymphocytes, the categories of the lymphocyte spots 1 and 3 are corrected to be positive tumor cells, and the voting results of the lymphocyte spots 2 and 4 are less than the corresponding category confidence degrees, so that the categories of the lymphocyte spots 2 and 4 are not corrected, as shown in fig. 8. If the discrimination result of the computer-aided discrimination does not contain the category confidence, setting the standard confidence VtWhen the result of voting for lymphocyte spots 1-4 is greater than the standard confidence V, 0.8tThat is, all of the lymphocyte spots 1-4 were corrected to positive tumor cells, as shown in FIG. 9, with a standard confidence VtAccording to the experience of the pathologist.
Exemplary computer program product and computer-readable storage Medium
In addition to the above-described methods, embodiments of the present application may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform the steps in the method for converting a staining style of a pathological section according to various embodiments of the present application described in the "exemplary methods" section of this specification above.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform steps in a method for improving the accuracy of the index of the pathology image KI67 according to various embodiments of the present application described in the "exemplary methods" section above in this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application. The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (6)

1. A method for improving the accuracy of an index of a pathological image KI67 is characterized by comprising the following steps:
s10, recognizing lymphocytes and positive tumor cells on the pathological section under the current visual field through a computer-aided discrimination algorithm, wherein the total number of the lymphocytes is n, and the total number of the positive tumor cells is m;
s20, screening central points of all the lymphocytes and positive tumor cells;
s30, extracting information of each lymphocyte and positive tumor cell, wherein the information comprises the abscissa and the ordinate of the cell and the average value of the red channel pixel intensity on a circular pixel area taking the central point as the center;
s40, calculating the pixel distance between the ith lymphocyte and the jth positive tumor cell according to the information
Figure FDA0002874890950000011
Calculating the pixel intensity distance of the red channel between the ith lymphocyte and the jth positive tumor cell according to the information
Figure FDA0002874890950000012
Wherein i is more than or equal to 1 and less than or equal to n, and j is more than or equal to 1 and less than or equal to m;
s50, according to the pixel distance
Figure FDA0002874890950000013
And pixel intensity distance
Figure FDA0002874890950000014
Calculating the combined distance between the ith lymphocyte and the jth positive tumor cell
Figure FDA0002874890950000015
Generating a combinatorial matrix Mcombine，McombineThe ith row and the jth column of (A) elements are
Figure FDA0002874890950000016
S60, combining the matrix McombineAdding the values of the second dimension to obtain an n-dimensional vector VvoteI.e. the current view imageVoting results of all positive tumor cells in the tumor cell against n lymphocytes;
s70, judging whether the voting result is effective, and if the voting result is effective, correcting the lymphocyte misjudged as the positive tumor cell into the positive tumor cell;
and S80, calculating a KI67 index according to the corrected number of the positive tumor cells.
2. The method for improving the accuracy of the KI67 indicator of the pathological image according to claim 1, wherein the S70 specifically includes:
judging whether the discrimination result of the computer-aided discrimination contains a category confidence level,
if yes, obtaining the confidence coefficient of the categories of the m lymphocytes, and marking as VptJudging the n lymphocytes one by one, and for the ith lymphocyte, if so
Figure FDA0002874890950000017
The voting result is considered to be valid, and the i-th lymphocyte is a positive tumor cell misjudged as a lymphocyte, so that the classification of the i-th lymphocyte is corrected to be a positive tumor cell, and if so, the classification is judged to be valid
Figure FDA0002874890950000018
The voting result is considered to be invalid;
if not, setting a standard confidence coefficient Vt，VtAs a constant, judging n lymphocytes one by one, and for the ith lymphocyte, if
Figure FDA0002874890950000019
The voting result is considered to be valid, and the i-th lymphocyte is a positive tumor cell misjudged as a lymphocyte, so that the classification of the i-th lymphocyte is corrected to be a positive tumor cell, and if so, the classification is judged to be valid
Figure FDA00028748909500000110
The voting result is considered invalid.
3. The method for improving the accuracy of the KI67 index of the pathological image according to claim 1, wherein the KI67 index is used for the pathological image
Figure FDA0002874890950000021
Expressed as the combined distance between the ith lymphocyte and the jth positive tumor cell, where α, β, dstd1And is dstd2All of which are constants, alpha is used to adjust the magnitude of the resulting value, and beta is used to adjust the relative weights of the pixel distance and the pixel intensity distance.
4. The method for improving the accuracy of the KI67 indicator of the pathological image according to claim 1, wherein the computer-aided discrimination algorithm includes a morphology-based conventional image algorithm and a deep learning algorithm.
5. The method for improving the accuracy of the KI67 index of the pathological image according to claim 1, wherein the lymphocytes and the positive tumor cells are all cell types output by the computer-aided identification algorithm.
6. A computer-readable medium, on which computer program instructions are stored, which, when executed by a processor, cause the processor to carry out the method for improving the accuracy of the KI67 indicator of a pathology image according to any one of claims 1 to 5.
CN202011615824.1A 2020-12-31 2020-12-31 Method for improving index accuracy of pathological image KI67 Active CN112801939B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011615824.1A CN112801939B (en) 2020-12-31 2020-12-31 Method for improving index accuracy of pathological image KI67

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011615824.1A CN112801939B (en) 2020-12-31 2020-12-31 Method for improving index accuracy of pathological image KI67

Publications (2)

Publication Number Publication Date
CN112801939A true CN112801939A (en) 2021-05-14
CN112801939B CN112801939B (en) 2022-07-22

Family

ID=75804680

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011615824.1A Active CN112801939B (en) 2020-12-31 2020-12-31 Method for improving index accuracy of pathological image KI67

Country Status (1)

Country Link
CN (1) CN112801939B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101799926A (en) * 2010-05-05 2010-08-11 福州大学 Automatically quantitative analysis system of Ki-67 immune-histochemical pathological image
US20150347702A1 (en) * 2012-12-28 2015-12-03 Ventana Medical Systems, Inc. Image Analysis for Breast Cancer Prognosis
CN106570505A (en) * 2016-11-01 2017-04-19 北京昆仑医云科技有限公司 Method for analyzing histopathologic image and system thereof
CN109196554A (en) * 2016-05-18 2019-01-11 豪夫迈·罗氏有限公司 Tumour measures of closeness
CN109906469A (en) * 2016-11-10 2019-06-18 豪夫迈·罗氏有限公司 Staging based on distance
WO2020014477A1 (en) * 2018-07-11 2020-01-16 The University Of North Carolina At Chapel Hill Methods, systems, and computer readable media for image analysis with deep learning to predict breast cancer classes
CN110705565A (en) * 2019-09-09 2020-01-17 西安电子科技大学 Lymph node tumor region identification method and device
CN110763678A (en) * 2019-09-12 2020-02-07 杭州迪英加科技有限公司 Pathological section interpretation method and system
CN111417958A (en) * 2017-12-07 2020-07-14 文塔纳医疗系统公司 Deep learning system and method for joint cell and region classification in biological images
CN111542830A (en) * 2017-12-29 2020-08-14 徕卡生物系统成像股份有限公司 Processing histological images using convolutional neural networks to identify tumors
CN111882561A (en) * 2020-06-18 2020-11-03 桂林电子科技大学 Cancer cell identification and diagnosis system

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101799926A (en) * 2010-05-05 2010-08-11 福州大学 Automatically quantitative analysis system of Ki-67 immune-histochemical pathological image
US20150347702A1 (en) * 2012-12-28 2015-12-03 Ventana Medical Systems, Inc. Image Analysis for Breast Cancer Prognosis
CN109196554A (en) * 2016-05-18 2019-01-11 豪夫迈·罗氏有限公司 Tumour measures of closeness
CN106570505A (en) * 2016-11-01 2017-04-19 北京昆仑医云科技有限公司 Method for analyzing histopathologic image and system thereof
CN109906469A (en) * 2016-11-10 2019-06-18 豪夫迈·罗氏有限公司 Staging based on distance
CN111417958A (en) * 2017-12-07 2020-07-14 文塔纳医疗系统公司 Deep learning system and method for joint cell and region classification in biological images
US20200342597A1 (en) * 2017-12-07 2020-10-29 Ventana Medical Systems, Inc. Deep-learning systems and methods for joint cell and region classification in biological images
CN111542830A (en) * 2017-12-29 2020-08-14 徕卡生物系统成像股份有限公司 Processing histological images using convolutional neural networks to identify tumors
WO2020014477A1 (en) * 2018-07-11 2020-01-16 The University Of North Carolina At Chapel Hill Methods, systems, and computer readable media for image analysis with deep learning to predict breast cancer classes
CN110705565A (en) * 2019-09-09 2020-01-17 西安电子科技大学 Lymph node tumor region identification method and device
CN110763678A (en) * 2019-09-12 2020-02-07 杭州迪英加科技有限公司 Pathological section interpretation method and system
CN111882561A (en) * 2020-06-18 2020-11-03 桂林电子科技大学 Cancer cell identification and diagnosis system

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
FUYONG XING, ET.AL: "Pixel-to-Pixel Learning With Weak Supervision for Single-Stage Nucleus Recognition in Ki67 Images", 《IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING》 *
HASSEN SEDDIK, ET.AL: "Identifying and classifying cancerous cells based on the Ki67 detector", 《2016 INTERNATIONAL IMAGE PROCESSING, APPLICATIONS AND SYSTEMS (IPAS)》 *
朱梅刚等: "《淋巴组织增生性病变良恶性鉴别诊断》", 31 October 2012, 广东科技出版社 *
王娇: "P16、Ki67在宫颈脱落细胞中的诊断意义及在宫颈病变筛查中的价值研究", 《中国优秀博硕士学位论文全文数据库(硕士)》 *
王林伟: "量子点多光谱成像结合计算机图像识别技术建立乳腺癌Ki67评估体系", 《中国博士学位论文全文数据库 (医药卫生科技辑)》 *
钟庄龙: "保留并修正中鼻甲的鼻内镜下鼻息肉手术后疗效评估和囊泡组织中Ki67与VEGF的表达及意义", 《中国优秀博硕士学位论文全文数据库(硕士)》 *

Also Published As

Publication number Publication date
CN112801939B (en) 2022-07-22

Similar Documents

Publication Publication Date Title
US12100146B2 (en) Assessing risk of breast cancer recurrence
US9558550B2 (en) Method and system for the automatic analysis of an image of a biological sample
US8369600B2 (en) Method and apparatus for detecting irregularities in tissue microarrays
US9122955B2 (en) Method and system of classifying medical images
CN109934278B (en) High-dimensionality feature selection method for information gain mixed neighborhood rough set
CN112507963B (en) Automatic generation of batch mask face samples and mask face recognition method
CN111462102B (en) Intelligent analysis system and method based on novel coronavirus pneumonia X-ray chest radiography
EP3611695A1 (en) Generating annotation data of tissue images
Xu et al. Using transfer learning on whole slide images to predict tumor mutational burden in bladder cancer patients
CN113951834A (en) Alzheimer disease classification prediction method based on visual Transformer algorithm
KR20180045473A (en) System, method and computer program for melanoma detection using image analysis
US20150199587A1 (en) Image processing device, image processing method, and image processing program
CA3204439A1 (en) Quantification of conditions on biomedical images across staining modalities using a multi-task deep learning framework
CN112884725B (en) Correction method for neural network model output result for cell discrimination
CN112801939B (en) Method for improving index accuracy of pathological image KI67
CN116309465B (en) Tongue image detection and positioning method based on improved YOLOv5 in natural environment
CN115775410B (en) Eye image data processing method, system, storage medium and electronic equipment
CN115861598A (en) Polyp detection method based on convolutional neural network
CN109800820A (en) A kind of classification method based on ultrasonic contrast image uniform degree
CN115331731A (en) Copy number variation detection method, device, equipment and computer readable medium
CN111368872A (en) Breast cancer mitosis cell detection method based on fusion characteristics and verification model
US20220318999A1 (en) Deep learning based blob detection systems and methods
Sarmiento Vega et al. A Method for Unsupervised Semi-Quantification of Inmunohistochemical Staining with Beta Divergences
EP4113439A1 (en) Determining a location at which a given feature is represented in medical imaging data
WO2024095319A1 (en) Information processing apparatus, information processing method, and recording medium

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
GR01 Patent grant
GR01 Patent grant