WO2024011400A1 - 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒 - Google Patents

生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒 Download PDF

Info

Publication number
WO2024011400A1
WO2024011400A1 PCT/CN2022/105113 CN2022105113W WO2024011400A1 WO 2024011400 A1 WO2024011400 A1 WO 2024011400A1 CN 2022105113 W CN2022105113 W CN 2022105113W WO 2024011400 A1 WO2024011400 A1 WO 2024011400A1
Authority
WO
WIPO (PCT)
Prior art keywords
breast cancer
tumor cells
invasive breast
ready
antibody
Prior art date
Application number
PCT/CN2022/105113
Other languages
English (en)
French (fr)
Inventor
杨清海
陈惠玲
程本亮
周洪辉
吴楠楠
Original Assignee
福州迈新生物技术开发有限公司
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 福州迈新生物技术开发有限公司 filed Critical 福州迈新生物技术开发有限公司
Priority to PCT/CN2022/105113 priority Critical patent/WO2024011400A1/zh
Publication of WO2024011400A1 publication Critical patent/WO2024011400A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N1/00—Sampling; Preparing specimens for investigation
    • G01N1/28—Preparing specimens for investigation including physical details of (bio-)chemical methods covered elsewhere, e.g. G01N33/50, C12Q
    • G01N1/30—Staining; Impregnating ; Fixation; Dehydration; Multistep processes for preparing samples of tissue, cell or nucleic acid material and the like for analysis

Definitions

  • the present application relates to the field of computer technology, and in particular to methods, storage devices and kits for generating training data for calculating Ki-67 counts of invasive breast cancer cells.
  • Ki-67 is a type of proliferating cell nuclear antigen related to the cell cycle and is currently one of the most widely used cell proliferation markers. Studies have shown that the positive expression of Ki-67 can be used as an important indicator to judge the biological behavior of breast cancer, and the expression level is related to the stage of breast cancer and lymph node metastasis. It is also an important indicator in current research on cell proliferation activity. Studies have shown that Ki-67 can guide the prognostic factors of breast cancer and guide the treatment of breast cancer. Ki-67 is of very important clinical significance.
  • the "2021 CSCO Breast Cancer Diagnosis and Treatment Guidelines" stipulates that all cases of breast invasive cancer should be tested for Ki-67, and the percentage of positively stained cells in the nucleus of cancer cells should be reported.
  • Ki-67 The percentage of positive tumor cells was skewed.
  • the interpretation results of Ki-67 strictly rely on the clinical experience of clinical pathologists. The results are all estimates and are greatly affected by the subjective influence of clinical pathologists.
  • this application provides a method for generating training data for calculating Ki-67 counts of invasive breast cancer cells to solve the problem of deviation in the percentage of Ki-67-positive tumor cells caused by the existing technology, Ki
  • the interpretation results of -67 strictly rely on the clinical experience of clinical pathologists, and the results are all estimated readings, which are greatly affected by the subjective influence of clinical pathologists and other technical issues.
  • the specific technical solutions are as follows:
  • a method of generating training data for calculating Ki-67 counts in invasive breast cancer cells comprising:
  • the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • using a specific combination of immunohistochemical markers to perform immunohistochemical staining on the target object through a preset method includes the steps:
  • the first preprocessing operation on paraffin sections includes the steps:
  • Paraffin sections are dewaxed, hydrated, and rinsed with running water;
  • adding ready-to-use combined primary antibodies and ready-to-use combined secondary antibodies to the pretreated paraffin sections respectively includes the steps:
  • PBS rinses the first preset number of times
  • the second pretreatment operation is performed on the paraffin sections after adding the ready-to-use combined primary antibody and the ready-to-use combined secondary antibody, including the steps of: adding chromogenic solution A; adding chromogenic solution B; counterstaining; dehydration , transparency and sealing.
  • the ready-to-use combination primary antibody includes: myoepithelial marker antibody, glandular epithelial marker antibody and Ki-67 antibody; the myoepithelial marker antibody and the glandular epithelial marker antibody are located in the cytoplasm and/or Or cell nucleus; the ready-to-use combination secondary antibodies include: ready-to-use enzyme-labeled combination secondary antibodies.
  • myoepithelial marker antibodies include: CK5;
  • the glandular epithelial marker antibodies include: CK8/18;
  • the Ki-67 antibody is a rabbit monoclonal antibody
  • the glandular epithelial marker antibody is a CK8/18 mouse monoclonal antibody
  • the myoepithelial marker antibody is a CK5 rabbit monoclonal antibody.
  • the invasive breast cancer area and the in-situ cancer area are taken as the data set for training the AI to distinguish the invasive breast cancer and the in-situ cancer, and the tumor cells in the invasive breast cancer area are selected as the training AI to calculate the invasiveness
  • Positive tumor cells are used as a data set for training AI to calculate the number of Ki-67 negative tumor cells.
  • the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • the digital pathology image is obtained in the following manner.
  • a kit for generating training data for calculating Ki-67 counts of invasive breast cancer cells is also provided.
  • the specific technical solution is as follows:
  • kits for generating training data for calculating Ki-67 counts of invasive breast cancer cells includes: a ready-to-use combined primary antibody;
  • the ready-to-use combination primary antibodies include: myoepithelial marker antibodies, glandular epithelial marker antibodies and Ki-67 antibodies;
  • the myoepithelial marker antibody and the glandular epithelial marker antibody are located in the cytoplasm and/or nucleus.
  • the beneficial effects of the present invention are: a method for generating training data for calculating Ki-67 counts of invasive breast cancer cells, including: performing immunohistochemistry on a target subject using a specific combination of immunohistochemical markers through a preset method Staining; select the invasive breast cancer area and the in-situ cancer area as the data set for training AI to distinguish invasive breast cancer and in-situ cancer; select the tumor cells in the invasive breast cancer area as the training AI to calculate the total number of invasive breast cancer tumor cells Data set; select Ki-67 positive tumor cells in the invasive breast cancer area as the data set for training AI to calculate the number of Ki-67 positive tumor cells; select non-Ki-67 positive tumor cells in the invasive breast cancer area as the training AI to calculate Ki A data set of the number of Ki-67 negative tumor cells; the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • the above method can accurately distinguish invasive cancer and carcinoma in situ as well as
  • Figure 1 is a flow chart of a method for generating training data for calculating Ki-67 counts of invasive breast cancer cells according to the specific embodiment
  • Figure 2 is a flow chart for performing immunohistochemical staining on a target object using a specific combination of immunohistochemical markers using a preset method according to the specific embodiment
  • Figure 3 is a flow chart of the first preprocessing operation on paraffin sections according to the specific embodiment
  • Figure 4 is a flow chart for adding ready-to-use combined primary antibodies and ready-to-use combined secondary antibodies to pre-treated paraffin sections according to the specific embodiment
  • Figure 5 is a module schematic diagram of a storage device according to the specific embodiment
  • Figure 6 is a schematic diagram of the components of the reagent materials described in the specific embodiment
  • Figure 7 is a schematic diagram of the immunohistochemical staining results of CK5 and p63/p40 in breast myoepithelial cells according to the specific embodiment
  • Figure 8 is a schematic diagram of the immunohistochemical staining results of CK5 and Calponin in breast myoepithelial cells according to the specific embodiment
  • Figure 9 is a schematic diagram of the staining results of CK8/18+CK5 and CK(pan)+CK5 in 90 cases of breast cancer according to the specific embodiment
  • Figure 10a is a schematic diagram of the dyeing results of the combination according to the specific embodiment
  • Figure 10b is a schematic diagram of the combined two staining results described in the specific embodiment
  • Figure 11a is a schematic diagram of the first sequential dyeing result described in the specific embodiment
  • Figure 11b is a schematic diagram of the second sequential staining result described in the specific embodiment.
  • Figure 12 is a schematic diagram of immunohistochemical staining of a target object using a specific combination of immunohistochemical markers using a preset method as described in the specific embodiment.
  • an embodiment means that a particular feature, structure or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application.
  • the word “embodiment” appearing in various places in the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or correlation with other embodiments. In principle, in this application, as long as there is no technical contradiction or conflict, the technical features mentioned in each embodiment can be combined in any way to form a corresponding implementable technical solution.
  • a method of generating training data for calculating Ki-67 counts of invasive breast cancer cells includes steps S101 to S106.
  • step S101 a specific combination of immunohistochemical markers is used to perform immunohistochemical staining on the target object through a preset method.
  • step S102 the invasive breast cancer area and the carcinoma in situ area are taken as a data set for training the AI to distinguish between invasive breast cancer and carcinoma in situ.
  • step S103 tumor cells in the invasive breast cancer area are selected as a data set for training the AI to calculate the total number of invasive breast cancer tumor cells.
  • Ki-67-positive tumor cells in the invasive breast cancer area are selected as a data set for training the AI to calculate the number of Ki-67-positive tumor cells.
  • step S105 non-Ki-67-positive tumor cells in the invasive breast cancer area are selected as a data set for training the AI to calculate the number of Ki-67-negative tumor cells.
  • step S106 the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • FIG. 2 shows that step S101 specifically also includes steps S201 to S203.
  • step S201 a first preprocessing operation is performed on the paraffin sections.
  • step S202 a ready-to-use combined primary antibody and a ready-to-use combined secondary antibody are added to the pre-processed paraffin sections respectively.
  • step S203 a second preprocessing operation is performed on the paraffin sections after adding the ready-to-use combined primary antibody and the ready-to-use combined secondary antibody.
  • FIG. 3 shows that step S201 specifically also includes steps S301 to S303.
  • step S301 the paraffin sections are dewaxed, hydrated, and rinsed with running water.
  • step S302 antigen retrieval is performed on the paraffin sections washed with running water.
  • step S303 blocking endogenous peroxidase is added to the paraffin sections after antigen repair.
  • steps S301 to S303 can also be as follows:
  • reagent 1 can be the special antigen retrieval solution (50x) for the double-staining kit), and dilute it with distilled water at a ratio of 1:50.
  • Chromogenic solution A needs to be temporarily prepared before use. If there is a little precipitation during preparation, it can be filtered before use. It will not affect the dyeing effect and quality.
  • Preparation method Mix buffer and chromogen in a ratio of 20:1. The prepared chromogenic solution A should be stored in a dark environment at 2-8°C and will be valid within 7 days.
  • Chromogenic solution B needs to be prepared temporarily before use. Preparation method: Mix buffer and chromogen in a ratio of 1:1. The prepared chromogenic solution B should be stored away from light and will be effective within 30 minutes.
  • FIG. 4 shows that step S202 specifically also includes steps S401 to S404.
  • step S401 remove the PBS solution, add the ready-to-use combined primary antibody, and incubate at room temperature for a first preset time.
  • step S402 PBS rinses the first preset number of times.
  • step S403 remove the PBS solution, add the ready-to-use combined second antibody, and incubate at room temperature for a second preset time.
  • step S404 PBS is rinsed for a second preset number of times.
  • steps S401 to S404 can be as follows:
  • Step S203 specifically includes the following steps:
  • hematoxylin staining solution (reagent 7) for light staining for 10 to 30 seconds, and rinse with running water;
  • the acquired slide is scanned to generate a digital pathology image, and then based on the tumor cells in the invasive breast cancer area (the cytoplasm is red and the nucleus is blue or brown), it is used as a data set for training AI to calculate the total number of invasive breast cancer tumor cells.
  • Use Ki-67-positive tumor cells (cells with red cytoplasm and brown nuclei) in the invasive breast cancer area as a data set for training AI to calculate the number of Ki-67-positive tumor cells, and select non-Ki-67-positive tumor cells in the invasive breast cancer area.
  • 67 positive tumor cells are used as a data set for training AI to calculate the number of Ki-67 negative tumor cells.
  • the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • the above method can accurately distinguish invasive cancer and carcinoma in situ as well as stromal and lymphocyte areas, and calculate the percentage of Ki-67 positive tumor cells in invasive cancer cells in the invasive cancer area.
  • the ready-to-use combination primary antibody includes: myoepithelial marker antibody, glandular epithelial marker antibody and Ki-67 antibody; the myoepithelial marker antibody and the glandular epithelial marker antibody are positioned at Cytoplasm and/or cell nucleus; the ready-to-use combination secondary antibodies include: ready-to-use enzyme-labeled combination secondary antibodies.
  • the myoepithelial marker antibodies may also include: CK5, Calponin, and/or p63/p40.
  • CK5 Calponin
  • p63/p40 a myoepithelial marker
  • the results showed that compared with CK5, P63/p40 exists as a myoepithelial marker and the continuity of myoepithelial staining in breast cancer tissue is poor.
  • Calponin stains smooth muscle and fibroblasts.
  • the staining results can easily make it difficult to distinguish between carcinoma in situ, minimally invasive carcinoma and invasive carcinoma. Therefore, the best myoepithelial marker for breast cancer is CK5.
  • the myoepithelial marker antibody is: CK5.
  • the immunohistochemical staining results of CK5 and p63/p40 in breast myoepithelial cells are shown in Figure 7.
  • the immunohistochemical staining results of CK5 and Calponin in breast myoepithelial cells are shown in Figure 8.
  • the glandular epithelial marker antibodies may include: CK8/18, CK(pan). Immunohistochemical double staining of breast cancer tissue was performed using CK8/18+CK5 and CK(pan)+CK5 respectively. The results showed that the two antibody combinations showed no significant difference in the 90 cases of breast cancer tested, but in the breast cancer tissue Due to the superimposed staining of CK (pan) and CK5 on the myoepithelium, it can be seen that the color and luster of myoepithelium are slightly different from that of CK8/18+CK5 double staining. Therefore, in some embodiments, preferably, the glandular epithelial marker antibody is CK8 /18. The staining results of CK8/18+CK5 and CK(pan)+CK5 are shown in Figure 9.
  • Figure 10a is a schematic diagram of the staining results of combination one
  • Figure 10b is a schematic diagram of the staining results of combination two.
  • the first sequence is to develop HRP first, and then to develop AP.
  • the schematic diagram of the staining results is shown in Figure 11a; the second sequence is to develop AP first, Then the HRP was colored, and the schematic diagram of the staining results is shown in Figure 11b.
  • the results showed that among the 90 samples compared, the positive rate of Ki-67 (brown), which develops color first, was significantly higher than that of color later.
  • the Ki-67 antibody is a rabbit monoclonal antibody
  • the glandular epithelial marker antibody is a CK8/18 mouse monoclonal antibody
  • the myoepithelial marker antibody is a CK5 rabbit monoclonal antibody.
  • the invasive breast cancer area and the in-situ cancer area were selected as the data set for training AI to distinguish invasive breast cancer and in-situ cancer, and the tumor cells in the invasive breast cancer area were selected as the training AI to calculate invasive breast cancer.
  • the data set of the total number of tumor cells, and the selection of Ki-67 positive tumor cells in the invasive breast cancer area as the data set for training the AI to calculate the number of Ki-67 positive tumor cells, and the selection of non-Ki-67 positive tumors in the invasive breast cancer area Cells are used as a data set for training AI to calculate the number of Ki-67 negative tumor cells.
  • the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • the digital pathology image is obtained in the following manner.
  • reagent 1 can be the special antigen retrieval solution (50x) for the double-staining kit), and dilute it with distilled water at a ratio of 1:50.
  • Chromogenic solution A needs to be temporarily prepared before use. If there is a little precipitation during preparation, it can be filtered before use. It will not affect the dyeing effect and quality.
  • Preparation method Mix buffer and chromogen in a ratio of 20:1. The prepared chromogenic solution A should be stored in a dark environment at 2-8°C and will be valid within 7 days.
  • Chromogenic solution B needs to be prepared temporarily before use. Preparation method: Mix buffer and chromogen in a ratio of 1:1. The prepared chromogenic solution B should be stored away from light and will be effective within 30 minutes.
  • Step S203 specifically includes the following steps:
  • hematoxylin staining solution (reagent 7) for light staining for 10 to 30 seconds, and rinse with running water;
  • the acquired slide is scanned to generate a digital pathology image, and then based on the tumor cells in the invasive breast cancer area (the cytoplasm is red and the nucleus is blue or brown), it is used as a data set for training AI to calculate the total number of invasive breast cancer tumor cells.
  • Use Ki-67-positive tumor cells (cells with red cytoplasm and brown nuclei) in the invasive breast cancer area as a data set for training AI to calculate the number of Ki-67-positive tumor cells, and select non-Ki-67-positive tumor cells in the invasive breast cancer area.
  • 67 positive tumor cells are used as a data set for training AI to calculate the number of Ki-67 negative tumor cells.
  • the total number of tumor cells in the invasive breast cancer area is equal to the sum of Ki-67 positive tumor cells and Ki-67 negative tumor cells.
  • the above-mentioned storage device 500 can accurately distinguish invasive cancer and carcinoma in situ as well as stromal and lymphocyte areas, and calculate the percentage of Ki-67 positive tumor cells in the invasive cancer cells in the invasive cancer area.
  • the ready-to-use combination primary antibody includes: myoepithelial marker antibody, glandular epithelial marker antibody and Ki-67 antibody; the myoepithelial marker antibody and the glandular epithelial marker antibody are positioned at Cytoplasm and/or cell nucleus; the ready-to-use combination secondary antibodies include: ready-to-use enzyme-labeled combination secondary antibodies.
  • the myoepithelial marker antibodies may also include: CK5, Calponin, and/or p63/p40.
  • CK5 Calponin
  • p63/p40 a myoepithelial marker
  • the results showed that compared with CK5, P63/p40 exists as a myoepithelial marker and the continuity of myoepithelial staining in breast cancer tissue is poor.
  • Calponin stains smooth muscle and fibroblasts.
  • the staining results can easily make it difficult to distinguish between carcinoma in situ, minimally invasive carcinoma and invasive carcinoma. Therefore, the best myoepithelial marker for breast cancer is CK5.
  • the myoepithelial marker antibody is: CK5.
  • the immunohistochemical staining results of CK5 and p63/p40 in breast myoepithelial cells are shown in Figure 7.
  • the immunohistochemical staining results of CK5 and Calponin in breast myoepithelial cells are shown in Figure 8.
  • the glandular epithelial marker antibodies may include: CK8/18, CK(pan). Immunohistochemical double staining of breast cancer tissue was performed using CK8/18+CK5 and CK(pan)+CK5 respectively. The results showed that the two antibody combinations showed no significant difference in the 90 cases of breast cancer tested, but in the breast cancer tissue Due to the superimposed staining of CK (pan) and CK5 on the myoepithelium, it can be seen that the color and luster of myoepithelium are slightly different from that of CK8/18+CK5 double staining. Therefore, in some embodiments, preferably, the glandular epithelial marker antibody is CK8 /18. The staining results of CK8/18+CK5 and CK(pan)+CK5 are shown in Figure 9.
  • Figure 10a is a schematic diagram of the staining results of combination one
  • Figure 10b is a schematic diagram of the staining results of combination two.
  • the first sequence is to develop HRP first, and then to develop AP.
  • the schematic diagram of the staining results is shown in Figure 11a; the second sequence is to develop AP first, Then the HRP was colored, and the schematic diagram of the staining results is shown in Figure 11b.
  • the results showed that among the 90 samples compared, the positive rate of Ki-67 (brown), which develops color first, was significantly higher than that of color later.
  • the Ki-67 antibody uses a rabbit monoclonal antibody cloned as MXR002, and the glandular epithelial marker antibody uses a mouse monoclonal antibody CK8/18 cloned as MX004+MX035, so The myoepithelial marker antibody used was rabbit monoclonal antibody CK5 cloned as EP24.
  • the kit includes: a ready-to-use combined primary antibody; the ready-to-use combined primary antibody includes: myoepithelial marker antibody, glandular epithelial marker antibody and Ki-67 antibody; the myoepithelial marker antibody and the Antibodies to glandular epithelial markers localize to the cytoplasm and/or nucleus.
  • the myoepithelial marker antibodies may also include: CK5, Calponin, and/or p63/p40.
  • CK5 Calponin
  • p63/p40 a myoepithelial marker
  • the results showed that compared with CK5, P63/p40 exists as a myoepithelial marker and the continuity of myoepithelial staining in breast cancer tissue is poor.
  • Calponin stains smooth muscle and fibroblasts.
  • the staining results can easily make it difficult to distinguish between carcinoma in situ, minimally invasive carcinoma and invasive carcinoma. Therefore, the best myoepithelial marker for breast cancer is CK5.
  • the myoepithelial marker antibody is: CK5.
  • the immunohistochemical staining results of CK5 and p63/p40 in breast myoepithelial cells are shown in Figure 7.
  • the immunohistochemical staining results of CK5 and Calponin in breast myoepithelial cells are shown in Figure 8.
  • the glandular epithelial marker antibodies may include: CK8/18, CK(pan). Immunohistochemical double staining of breast cancer tissues was performed using CK8/18+CK5 and CK(pan) +CK5 respectively. The results showed that the two antibody combinations showed no significant difference in the 90 cases of breast cancer tested, but in the breast cancer tissues Due to the superimposed staining of CK (pan) and CK5 on the myoepithelium, it can be seen that the color and luster of myoepithelium are slightly different from that of CK8/18+CK5 double staining. Therefore, in some embodiments, preferably, the glandular epithelial marker antibody is CK8 /18. The staining results of CK8/18+CK5 and CK(pan)+CK5 are shown in Figure 9.
  • Figure 10a is a schematic diagram of the staining results of combination one
  • Figure 10b is a schematic diagram of the staining results of combination two.
  • the first sequence is to develop HRP first, and then to develop AP.
  • the schematic diagram of the staining results is shown in Figure 11a; the second sequence is to develop AP first, Then the HRP was colored, and the schematic diagram of the staining results is shown in Figure 11b.
  • the results showed that among the 90 samples compared, the positive rate of Ki-67 (brown), which develops color first, was significantly higher than that of color later.
  • the Ki-67 antibody is a rabbit monoclonal antibody
  • the glandular epithelial marker antibody is a CK8/18 mouse monoclonal antibody
  • the myoepithelial marker antibody is a CK5 rabbit monoclonal antibody.

Abstract

涉及计算机技术领域，特别涉及生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备（500）和试剂盒。生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，包括：通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色（S101）；选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集（S103）；选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集（S104）。通过该方法可准确区分浸润癌和原位癌以及间质和淋巴细胞区域，并在浸润癌区域计算Ki-67阳性肿瘤细胞占浸润癌细胞的百分比。

Description

生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒 技术领域
本申请涉及计算机技术领域，特别涉及生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒。
背景技术
Ki-67是一类与细胞周期相关的增殖细胞核抗原,并且是目前应用最广泛的细胞增殖标志之一。有研究表明Ki-67阳性表达可以作为判断乳腺癌生物学行为的重要指标,而且表达水平与乳腺癌的分期和淋巴结的转移等相关,也是目前细胞增殖活性研究的重要指标。有研究表明Ki-67可以指导乳腺癌的预后因素以及指导乳腺癌的治疗,在临床上Ki-67具有非常重要的意义。《2021CSCO乳腺癌诊疗指南》规定，应对所有乳腺浸润性癌病例进行Ki-67的检测，并对癌细胞核中阳性染色细胞所占的百分比进行报告，阳性定义为浸润癌细胞核任何程度的棕色染色。2021年“乳腺癌Ki-67国际工作组评估指南”推荐采用标准化的“打字机”视觉评估法进行判读，要有严格的质量评估保证和控制体系，确保分析的有效性。根据最新的评估标准，工作组明确将Ki-67＞30％定义为高表达，临床中，在面对HR阳性、HER-2阴性、T1N0的乳腺癌病人时，若Ki-67高表达，显然无需基因检测即可决定是否化疗。
在日常工作中，在H&E染色切片配合Ki-67单染进行半定量判读时，由于浸润癌和原位癌难以区分以及间质和淋巴细胞(也可呈Ki-67阳性)，可能导致Ki-67阳性肿瘤细胞所占的百分比出现偏差。Ki-67的判读结果严格依赖于临床病理医生的临床经验，结果均为估读，受临床病理医生的主观影响较大。
发明内容
鉴于上述问题，本申请提供了一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，用以解决现有技术导致Ki-67阳性肿瘤细胞所占的百分比出现偏差，Ki-67的判读结果严格依赖于临床病理医生的临床经验，结果均为估读，受临床病理医生的主观影响较大等技术问题。具体技术方案如下：
一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，包括：
通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色；
取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集；
选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集；
选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集；
选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集；
浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
进一步的，所述通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色，包括步骤：
对石蜡切片进行第一预处理操作；
对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗；
对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
进一步的，所述对石蜡切片进行第一预处理操作，包括步骤：
石蜡切片脱蜡、水化，流水冲洗；
对流水冲洗后的石蜡切片进行抗原修复；
对抗原修复后的石蜡切片添加阻断内源性过氧化物酶。
进一步的，所述对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗，包括步骤：
除去PBS液，加即用型组合一抗，室温下孵育第一预设时间；
PBS冲洗第一预设次数；
除去PBS液，加即用型组合第二抗体，室温下孵育第二预设时间；
PBS冲洗第二预设次数。
进一步的，所述对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作，包括步骤：添加显色液A；添加显色液B；复染；脱水、透明和封片。
进一步的，所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核；所述即用型组合二抗包括：即用型酶标组合二抗。
进一步的，所述肌上皮标记物抗体包括：CK5；
所述腺上皮标记物抗体包括：CK8/18；
所述Ki-67抗体采用兔单克隆抗体，所述腺上皮标记物抗体采用CK8/18鼠单克隆抗体，所述肌上皮标记物抗体采用CK5兔单克隆抗体。
为解决上述技术问题，还提供了一种存储设备，具体技术方案如下：
一种存储设备，其中存储有指令集，所述指令集用于执行：获取通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色后的数字病理图像；
在所述数字病理图像中取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集，及选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集，及选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集，及选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
进一步的，所述数字病理图像通过如下方式获取；
对石蜡切片进行第一预处理操作；
对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗；
对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
为解决上述技术问题，还提供了一种用于生成计算浸润性乳腺癌细胞Ki-67计数的训练数据的试剂盒，具体技术方案如下：
一种用于生成计算浸润性乳腺癌细胞Ki-67计数的训练数据的试剂盒，所述试剂盒包括：即用型组合一抗；
所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；
所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核。
本发明的有益效果是：一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，包括：通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色；取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集；选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集；选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集；选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集；浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。通过上述方法可准确区分浸润癌和原位癌以及间质和淋巴细胞区域，并在浸润癌区域计算Ki-67阳性肿瘤细胞占浸润癌细胞的百分比。
上述发明内容相关记载仅是本申请技术方案的概述，为了让本领域普通技术人员能够更清楚地了解本申请的技术方案，进而可以依据说明书的文字及附图记载的内容予以实施，并且为了让本申请的上述目的及其它目的、特征和优点能够更易于理解，以下结合本申请的具体实施方式及附图进行说明。
附图说明
附图仅用于示出本申请具体实施方式以及其他相关内容的原理、实现方式、应用、特点以及效果等，并不能认为是对本申请的限制。
在说明书附图中：
图1为具体实施方式所述一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法的流程图；
图2为具体实施方式所述通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色的流程图；
图3为具体实施方式所述对石蜡切片进行第一预处理操作的流程图；
图4为具体实施方式所述对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗的流程图；
图5为具体实施方式所述一种存储设备的模块示意图；
图6为具体实施方式所述试剂材料组分示意图；
图7为具体实施方式所述CK5与p63/p40在乳腺肌上皮细胞中的免疫组化染色结果示意图；
图8为具体实施方式所述CK5与Calponin在乳腺肌上皮细胞中的免疫组化染色结果示意图；
图9为具体实施方式所述CK8/18+CK5与CK(pan)+CK5在90例乳腺癌中的染色结果示意图；
图10a为具体实施方式所述组合一染色结果示意图；
图10b为具体实施方式所述组合二染色结果示意图；
图11a为具体实施方式所述第一种顺序染色结果示意图；
图11b为具体实施方式所述第二种顺序染色结果示意图；
图12为具体实施方式所述通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色后的示意图。
上述各附图中涉及的附图标记说明如下：
500、存储设备。
具体实施方式
为详细说明本申请可能的应用场景，技术原理，可实施的具体方案，能实现目的与效果等，以下结合所列举的具体实施例并配合附图详予说明。本文所记载的实施例仅用于更加清楚地说明本申请的技术方案，因此只作为示例，而不能以此来限制本申请的保护范围。
在本文中提及“实施例”意味着，结合实施例描述的特定特征、结构或特性可以包含在本申请的至少一个实施例中。在说明书中各个位置出现的“实施例”一词并不一定指代相同的实施例，亦不特别限定其与其它实施例之间的独立性或关联性。原则上，在本申请中，只要不存在技术矛盾或冲突，各实施例中所提到的各项技术特征均可以以任意方式进行组合，以形成相应的可实施的技术方案。
以下参阅图1至图4对一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法的具体实施方式展开说明。
如图1所示，一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法包括步骤S101至步骤S106。
在步骤S101中，通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色。
在步骤S102中，取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集。
在步骤S103中，选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集。
在步骤S104中，选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集。
在步骤S105中，选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集。
在步骤S106中，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
图2示出了步骤S101具体还包括了步骤S201至步骤S203。
在步骤S201中，对石蜡切片进行第一预处理操作。
在步骤S202中，对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗。
在步骤S203中，对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
图3示出了步骤S201具体还包括步骤S301至步骤S303。
在步骤S301中，石蜡切片脱蜡、水化，流水冲洗。
在步骤S302中，对流水冲洗后的石蜡切片进行抗原修复。
在步骤S303中，对抗原修复后的石蜡切片添加阻断内源性过氧化物酶。
其中步骤S301至步骤S303具体还可如下：
首先对以下染色工艺流程中需要的仪器、设备、温度条件及需预先进行配置的溶液做以下说明：
1、检测所需仪器、设备
移液器、油笔、电磁炉、计时器、孵育盒、染色架、盖玻片、不锈钢锅、光学显微镜、洗瓶。
2、溶液配制
2.1抗原修复液
根据所需抗原修复液的量，取试剂1(如图6所示，试剂1可为双染试剂盒专用抗原修复液(50x))，按1:50的比例加蒸馏水稀释。
2.2显色液A
显色液A需要在使用前临时配制，配制时如出现少许沉淀可过滤后使用，并不影响染色效果和质量。配制方法：将缓冲液与色原按20：1的比例混匀。配制后的显色液A在避光、2-8℃环境下存放，7天内有效。
2.3显色液B
显色液B需要在使用前临时配制。配制方法：将缓冲液与色原按1:1的比例混匀。配置好的显色液B避光存放，30分钟内有效。
3试验温度条件：室温
以下对步骤S301至步骤S303展开具体说明：
a)石蜡切片脱蜡、水化，流水冲洗。
b)抗原修复
不锈钢锅在电磁炉上大功率加热EDTA抗原修复液至沸腾；
将功率调至最小(处于保温状态)，将切片置于耐高温染片架上，放入已沸腾的修复液中，盖上锅盖，继续加热20分钟后，离开热源，自然冷却10分钟；
用自来水冲淋不锈钢锅外壁使之冷却，待锅中液体冷却至室温后取出切片；
蒸馏水冲洗3分钟×3次；
PBS溶液冲洗3分钟×3次。
c)阻断内源性过氧化物酶
除去PBS液，在油笔圈定的区域内加内源性过氧化物酶阻断剂(试剂2)，室温下孵育10分钟；
PBS冲洗3分钟×3次。
图4示出了步骤S202具体还包括步骤S401至步骤S404。
在步骤S401中，除去PBS液，加即用型组合一抗，室温下孵育第一预设时间。
在步骤S402中，PBS冲洗第一预设次数。
在步骤S403中，除去PBS液，加即用型组合第二抗体，室温下孵育第二预设时间。
在步骤S404中，PBS冲洗第二预设次数。
其中步骤S401至步骤S404具体可如下：
d)加即用型组合一抗
除去PBS液，加即用型组合一抗，室温下孵育60分钟；
PBS冲洗3分钟×3次。
e)加即用型组合第二抗体
除去PBS液，加即用型组合第二抗体，室温下孵育15分钟；
PBS冲洗3分钟×3次。
步骤S203具体还包括步骤：
f)加显色液A
除去PBS液，加配制好的显色液A，显色3～10分钟，必要时可在显微镜下观察掌握染色时间，阳性染色为棕黄(褐)色；
PBS冲洗3分钟×2次。
g)加显色液B
除去PBS液，加配制好的显色液B，显色8～20分钟，可在显微镜下观察掌握染色时间，阳性染色为红色；
流水冲洗终止显色。
h)复染
除去水分，加苏木素染色液(试剂7)浅染10～30秒，流水冲洗干净；
PBS冲洗返蓝，流水冲洗干净。
i)脱水、透明、封片
85％乙醇中，浸泡1分钟；
95％乙醇中，浸泡1分钟；
无水乙醇中，浸泡1分钟；
二甲苯中浸泡1分钟；
中性树胶和盖玻片封片。
对获取后的封片进行扫描生成数字病理图像，再根据浸润性乳腺癌区域内肿瘤细胞(细胞质呈红色，细胞核呈蓝色或棕色)作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集。以浸润性乳腺癌区域内Ki-67阳性肿瘤细胞(细胞质呈红色和细胞核呈棕色的细胞)作为训练AI计算Ki-67阳性肿瘤细胞数的数据集，及选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
如此通过上述方法可准确区分浸润癌和原位癌以及间质和淋巴细胞区域，并在浸润癌区域计算Ki-67阳性肿瘤细胞占浸润癌细胞的百分比。
正如上述所提及，所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核；所述即用型组合二抗包括：即用型酶标组合二抗。
在一些实施例中，所述肌上皮标记物抗体亦可包括：CK5、Calponin、和/或p63/p40。在通过使用CK5、Calponin、p63/p40分别对乳腺癌组织进行免疫组化单染，结果显示相较于CK5，P63/p40作为肌上皮标记物存在乳腺癌组织中的肌上皮染色连续性不佳的情况，而Calponin则存在平滑肌及纤维细胞染色，该染色结果容易造成原位癌、微浸润癌与浸润性癌的区分困难，因此，最佳的乳腺癌肌上皮标记物为CK5。故此，在一些实施例中，优选地，所述肌上皮标记物抗体为：CK5。其中CK5与p63/p40在 乳腺肌上皮细胞中的免疫组化染色结果如图7所示。CK5与Calponin在乳腺肌上皮细胞中的免疫组化染色结果如图8所示。
在一些实施例中，所述腺上皮标记物抗体可包括：CK8/18、CK(pan)。使用CK8/18+CK5和CK(pan)+CK5分别对乳腺癌组织进行免疫组化双染，结果显示两个抗体组合在测试的90例乳腺癌中未见明显差异，但在乳腺癌组织中的肌上皮上由于CK(pan)与CK5的叠加染色可见肌上皮显色色泽与CK8/18+CK5双染存在轻微差异，因此，在一些实施例中，优选地，腺上皮标记物抗体为CK8/18。其中CK8/18+CK5与CK(pan)+CK5的染色结果如图9所示。
此外在使用不同的检测系统组合对CK8/18、CK5和Ki-67进行显色，结果显示，组合一(CK8/18-AP(定位：细胞质，红色)，CK5-HRP(定位：肌上皮细胞细胞质，棕色)、Ki-67-HRP(定位：细胞核，棕色))相较于组合二(CK8/18-HRP(定位：细胞质，棕色)，CK5-AP(定位：肌上皮细胞细胞质，红色)、Ki-67-AP(定位：细胞核，红色))能够更显著地区分肿瘤细胞和Ki-67阳性肿瘤细胞。
其中图10a为组合一染色结果示意图，图10b为组合二染色结果示意图。
使用两种显色顺，即第一种顺序为先对HRP进行显色，而后再对AP进行显色，其染色结果示意图如图11a所示；第二种顺序为先对AP进行显色，而后再对HRP进行显色，其染色结果示意图如图11b所示。结果显示，在比对的90例样本中，均显示优先显色Ki-67(棕色)的阳性率明显高于后显色。
故此，在一些实施例中，优选地，所述Ki-67抗体采用兔单克隆抗体，所述腺上皮标记物抗体采用CK8/18鼠单克隆抗体，所述肌上皮标记物抗体采用CK5兔单克隆抗体。
最终的判读规则如下：
Figure PCTCN2022105113-appb-000001
其中最后的染色结果如图12所示。
以下参阅图5，对一种存储设备500的具体实施方式展开说明。
一种存储设备500，其中存储有指令集，所述指令集用于执行：获取通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色后的数字病理图像；在所述数字病理图像中取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集，及选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集，及选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集，及选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
进一步的，所述数字病理图像通过如下方式获取；
对石蜡切片进行第一预处理操作；
对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗；
对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
其具体如下：
首先对以下染色工艺流程中需要的仪器、设备、温度条件及需预先进行配置的溶液做以下说明：
1、检测所需仪器、设备
移液器、油笔、电磁炉、计时器、孵育盒、染色架、盖玻片、不锈钢锅、光学显微镜、洗瓶。
2、溶液配制
2.1抗原修复液
根据所需抗原修复液的量，取试剂1(如图6所示，试剂1可为双染试剂盒专用抗原修复液(50x))，按1:50的比例加蒸馏水稀释。
2.2显色液A
显色液A需要在使用前临时配制，配制时如出现少许沉淀可过滤后使用，并不影响染色效果和质量。配制方法：将缓冲液与色原按20：1的比例混匀。配制后的显色液A在避光、2-8℃环境下存放，7天内有效。
2.3显色液B
显色液B需要在使用前临时配制。配制方法：将缓冲液与色原按1:1的比例混匀。配置好的显色液B避光存放，30分钟内有效。
3试验温度条件：室温
a)石蜡切片脱蜡、水化，流水冲洗。
b)抗原修复
不锈钢锅在电磁炉上大功率加热EDTA抗原修复液至沸腾；
将功率调至最小(处于保温状态)，将切片置于耐高温染片架上，放入已沸腾的修复液中，盖上锅盖，继续加热20分钟后，离开热源，自然冷却10分钟；
用自来水冲淋不锈钢锅外壁使之冷却，待锅中液体冷却至室温后取出切片；
蒸馏水冲洗3分钟×3次；
PBS溶液冲洗3分钟×3次。
c)阻断内源性过氧化物酶
除去PBS液，在油笔圈定的区域内加内源性过氧化物酶阻断剂(试剂2)，室温下孵育10分钟；
PBS冲洗3分钟×3次。
d)加即用型组合一抗
除去PBS液，加即用型组合一抗，室温下孵育60分钟；
PBS冲洗3分钟×3次。
e)加即用型组合第二抗体
除去PBS液，加即用型组合第二抗体，室温下孵育15分钟；
PBS冲洗3分钟×3次。
步骤S203具体还包括步骤：
f)加显色液A
除去PBS液，加配制好的显色液A，显色3～10分钟，必要时可在显微镜下观察掌握染色时间，阳性染色为棕黄(褐)色；
PBS冲洗3分钟×2次。
g)加显色液B
除去PBS液，加配制好的显色液B，显色8～20分钟，可在显微镜下观察掌握染色时间，阳性染色为红色；
流水冲洗终止显色。
h)复染
除去水分，加苏木素染色液(试剂7)浅染10～30秒，流水冲洗干净；
PBS冲洗返蓝，流水冲洗干净。
i)脱水、透明、封片
85％乙醇中，浸泡1分钟；
95％乙醇中，浸泡1分钟；
无水乙醇中，浸泡1分钟；
二甲苯中浸泡1分钟；
中性树胶和盖玻片封片。
对获取后的封片进行扫描生成数字病理图像，再根据浸润性乳腺癌区域内肿瘤细胞(细胞质呈红色，细胞核呈蓝色或棕色)作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集。以浸润性乳腺癌区域内Ki-67阳性肿瘤细胞(细胞质呈红色和细胞核呈棕色的细胞)作为训练AI计算Ki-67阳性肿瘤细胞数的数据集，及选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
如此通过上述存储设备500可准确区分浸润癌和原位癌以及间质和淋巴细胞区域，并在浸润癌区域计算Ki-67阳性肿瘤细胞占浸润癌细胞的百分比。
正如上述所提及，所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核；所述即用型组合二抗包括：即用型酶标组合二抗。
在一些实施例中，所述肌上皮标记物抗体亦可包括：CK5、Calponin、和/或p63/p40。在通过使用CK5、Calponin、p63/p40分别对乳腺癌组织进行免疫组化单染，结果显示相较于CK5，P63/p40作为肌上皮标记物存在乳腺癌组织中的肌上皮染色连续性不佳的情况，而Calponin则存在平滑肌及纤维细胞染色，该染色结果容易造成原位癌、微浸润癌与浸润性癌的区分困难，因此，最佳的乳腺癌肌上皮标记物为CK5。故此，在一些实施例中，优选地，所述肌上皮标记物抗体为：CK5。其中CK5与p63/p40在乳腺肌上皮细胞中的免疫组化染色结果如图7所示。CK5与Calponin在乳腺肌上皮细胞中的免疫组化染色结果如图8所示。
在一些实施例中，所述腺上皮标记物抗体可包括：CK8/18、CK(pan)。使用CK8/18+CK5和CK(pan)+CK5分别对乳腺癌组织进行免疫组化双染，结果显示两个抗体组合在测试的90例乳腺癌中未见明显差异，但在乳腺癌组织中的肌上皮上由于CK(pan)与CK5的叠加染色可见肌上皮显色色泽与CK8/18+CK5双染存在轻微差异，因此，在一些实施例中，优选地，腺上皮标记物抗体为CK8/18。其中CK8/18+CK5与CK(pan)+CK5的染色结果如图9所示。
此外在使用不同的检测系统组合对CK8/18、CK5和Ki-67进行显色，结果显示，组合一(CK8/18-AP(定位：细胞质，红色)，CK5-HRP(定位：肌上皮细胞细胞质，棕色)、Ki-67-HRP(定位：细胞核，棕色))相较于组合二(CK8/18-HRP(定位：细胞质，棕色)，CK5-AP(定位：肌上皮细胞细胞质，红色)、Ki-67-AP(定位：细胞核，红色))能够更显著地区分肿瘤细胞和Ki-67阳性肿瘤细胞。
其中图10a为组合一染色结果示意图，图10b为组合二染色结果示意图。
使用两种显色顺，即第一种顺序为先对HRP进行显色，而后再对AP进行显色，其染色结果示意图如图11a所示；第二种顺序为先对AP进行显色，而后再对HRP进行显色，其染色结果示意图如图11b所示。结果显示，在比对的90例样本中，均显示优先显色Ki-67(棕色)的阳性率明显高于后显色。
故此，在一些实施例中，优选地，所述Ki-67抗体采用克隆为MXR002的兔单克隆抗体，所述腺上皮标记物抗体采用克隆为MX004+MX035的鼠单克隆抗体CK8/18，所述肌上皮标记物抗体采用克隆为EP24的兔单克隆抗体CK5。
最终的判读规则如下：
Figure PCTCN2022105113-appb-000002
其中最后的染色结果如图12所示。
以下对一种用于生成计算浸润性乳腺癌细胞Ki-67计数的训练数据的试剂盒的具体实施方式展开说明。
所述试剂盒包括：即用型组合一抗；所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核。
在一些实施例中，所述肌上皮标记物抗体亦可包括：CK5、Calponin、和/或p63/p40。在通过使用CK5、Calponin、p63/p40分别对乳腺癌组织进行免疫组化单染，结果显示相较于CK5，P63/p40作为肌上皮标记物存在乳腺癌组织中的肌上皮染色连续性不佳的情况，而Calponin则存在平滑肌及纤维细胞染色，该染色结果容易造成原位癌、微浸润癌与浸润性癌的区分困难，因此，最佳的乳腺癌肌上皮标记物为CK5。故此，在一些实施例中，优选地，所述肌上皮标记物抗体为：CK5。其中CK5与p63/p40在乳腺肌上皮细胞中的免疫组化染色结果如图7所示。CK5与Calponin在乳腺肌上皮细胞中的免疫组化染色结果如图8所示。
在一些实施例中，所述腺上皮标记物抗体可包括：CK8/18、CK(pan)。使用CK8/18+CK5和CK(pan) +CK5分别对乳腺癌组织进行免疫组化双染，结果显示两个抗体组合在测试的90例乳腺癌中未见明显差异，但在乳腺癌组织中的肌上皮上由于CK(pan)与CK5的叠加染色可见肌上皮显色色泽与CK8/18+CK5双染存在轻微差异，因此，在一些实施例中，优选地，腺上皮标记物抗体为CK8/18。其中CK8/18+CK5与CK(pan)+CK5的染色结果如图9所示。
此外在使用不同的检测系统组合对CK8/18、CK5和Ki-67进行显色，结果显示，组合一(CK8/18-AP(定位：细胞质，红色)，CK5-HRP(定位：肌上皮细胞细胞质，棕色)、Ki-67-HRP(定位：细胞核，棕色))相较于组合二(CK8/18-HRP(定位：细胞质，棕色)，CK5-AP(定位：肌上皮细胞细胞质，红色)、Ki-67-AP(定位：细胞核，红色))能够更显著地区分肿瘤细胞和Ki-67阳性肿瘤细胞。
其中图10a为组合一染色结果示意图，图10b为组合二染色结果示意图。
使用两种显色顺，即第一种顺序为先对HRP进行显色，而后再对AP进行显色，其染色结果示意图如图11a所示；第二种顺序为先对AP进行显色，而后再对HRP进行显色，其染色结果示意图如图11b所示。结果显示，在比对的90例样本中，均显示优先显色Ki-67(棕色)的阳性率明显高于后显色。
故此，在一些实施例中，优选地，所述Ki-67抗体采用兔单克隆抗体，所述腺上皮标记物抗体采用CK8/18鼠单克隆抗体，所述肌上皮标记物抗体采用CK5兔单克隆抗体。
最终的判读规则如下：
Figure PCTCN2022105113-appb-000003
其中最后的染色结果如图12所示。
最后需要说明的是，尽管在本申请的说明书文字及附图中已经对上述各实施例进行了描述，但并不能因此限制本申请的专利保护范围。凡是基于本申请的实质理念，利用本申请说明书文字及附图记载的内容所作的等效结构或等效流程替换或修改产生的技术方案，以及直接或间接地将以上实施例的技术方案实施于其他相关的技术领域等，均包括在本申请的专利保护范围之内。

Claims (10)

  1. 一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，包括：
    通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色；
    取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集；
    选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集；
    选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集；
    选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集；
    浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
  2. 根据权利要求1所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，所述通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色，包括步骤：
    对石蜡切片进行第一预处理操作；
    对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗；
    对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
  3. 根据权利要求2所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，所述对石蜡切片进行第一预处理操作，包括步骤：
    石蜡切片脱蜡、水化，流水冲洗；
    对流水冲洗后的石蜡切片进行抗原修复；
    对抗原修复后的石蜡切片添加阻断内源性过氧化物酶。
  4. 根据权利要求2所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，所述对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗，包括步骤：
    除去PBS液，加即用型组合一抗，室温下孵育第一预设时间；
    PBS冲洗第一预设次数；
    除去PBS液，加即用型组合第二抗体，室温下孵育第二预设时间；
    PBS冲洗第二预设次数。
  5. 根据权利要求2所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，所述对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作，包括步骤：
    添加显色液A；
    添加显色液B；
    复染；
    脱水、透明和封片。
  6. 根据权利要求2所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其特征在于，所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；
    所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核；
    所述即用型组合二抗包括：即用型酶标组合二抗。
  7. 根据权利要求6所述的一种生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法，其 特征在于，
    所述肌上皮标记物抗体包括：CK5；
    所述腺上皮标记物抗体包括：CK8/18；
    所述Ki-67抗体采用兔单克隆抗体，所述腺上皮标记物抗体采用CK8/18鼠单克隆抗体，所述肌上皮标记物抗体采用CK5兔单克隆抗体。
  8. 一种存储设备，其中存储有指令集，其特征在于，所述指令集用于执行：获取通过预设方法使用特定组合的免疫组化标志物对目标对象进行免疫组化染色后的数字病理图像；
    在所述数字病理图像中取浸润性乳腺癌区域和原位癌区域作为训练AI辨别浸润性乳腺癌和原位癌的数据集，及选取浸润性乳腺癌区域内肿瘤细胞作为训练AI计算浸润性乳腺癌肿瘤细胞总数的数据集，及选取浸润性乳腺癌区域内Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阳性肿瘤细胞数的数据集，及选取浸润性乳腺癌区域内非Ki-67阳性肿瘤细胞作为训练AI计算Ki-67阴性肿瘤细胞数的数据集，浸润性乳腺癌区域内肿瘤细胞总数等于Ki-67阳性肿瘤细胞与Ki-67阴性肿瘤细胞之和。
  9. 根据权利要求8所述的一种存储设备，其特征在于，所述数字病理图像通过如下方式获取；
    对石蜡切片进行第一预处理操作；
    对预处理后的石蜡切片分别添加即用型组合一抗和即用型组合二抗；
    对添加完即用型组合一抗和即用型组合二抗的石蜡切片进行第二预处理操作。
  10. 一种用于生成计算浸润性乳腺癌细胞Ki-67计数的训练数据的试剂盒，其特征在于，所述试剂盒包括：即用型组合一抗；
    所述即用型组合一抗包括：肌上皮标记物抗体、腺上皮标记物抗体和Ki-67抗体；
    所述肌上皮标记物抗体和所述腺上皮标记物抗体定位为细胞质和/或细胞核。
PCT/CN2022/105113 2022-07-12 2022-07-12 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒 WO2024011400A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/105113 WO2024011400A1 (zh) 2022-07-12 2022-07-12 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/105113 WO2024011400A1 (zh) 2022-07-12 2022-07-12 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒

Publications (1)

Publication Number Publication Date
WO2024011400A1 true WO2024011400A1 (zh) 2024-01-18

Family

ID=89535231

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/105113 WO2024011400A1 (zh) 2022-07-12 2022-07-12 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒

Country Status (1)

Country Link
WO (1) WO2024011400A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107330263A (zh) * 2017-06-26 2017-11-07 成都知识视觉科技有限公司 一种计算机辅助乳腺浸润性导管癌组织学分级的方法
CN111630604A (zh) * 2017-11-20 2020-09-04 纳诺全球公司 基于生物细胞或生物物质的检测的数据收集和分析
CN112465745A (zh) * 2020-10-26 2021-03-09 脉得智能科技(无锡)有限公司 一种基于全卷积回归网络的细胞计数方法
US20210150701A1 (en) * 2017-06-15 2021-05-20 Visiopharm A/S Method for training a deep learning model to obtain histopathological information from images
CN113011257A (zh) * 2021-02-08 2021-06-22 湖南省肿瘤医院 一种乳腺癌免疫组化人工智能判读方法
CN113762379A (zh) * 2021-09-07 2021-12-07 福州迈新生物技术开发有限公司 一种基于免疫组化生成训练数据的方法和存储设备

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20210150701A1 (en) * 2017-06-15 2021-05-20 Visiopharm A/S Method for training a deep learning model to obtain histopathological information from images
CN107330263A (zh) * 2017-06-26 2017-11-07 成都知识视觉科技有限公司 一种计算机辅助乳腺浸润性导管癌组织学分级的方法
CN111630604A (zh) * 2017-11-20 2020-09-04 纳诺全球公司 基于生物细胞或生物物质的检测的数据收集和分析
CN112465745A (zh) * 2020-10-26 2021-03-09 脉得智能科技(无锡)有限公司 一种基于全卷积回归网络的细胞计数方法
CN113011257A (zh) * 2021-02-08 2021-06-22 湖南省肿瘤医院 一种乳腺癌免疫组化人工智能判读方法
CN113762379A (zh) * 2021-09-07 2021-12-07 福州迈新生物技术开发有限公司 一种基于免疫组化生成训练数据的方法和存储设备

Similar Documents

Publication Publication Date Title
CN112326961B (zh) 一种非小细胞肺癌中pd-l1阳性肿瘤细胞比例的分析方法和存储设备
van den Brand et al. Sequential immunohistochemistry: a promising new tool for the pathology laboratory
Hatanaka et al. Cytometrical image analysis for immunohistochemical hormone receptor status in breast carcinomas
CN113762379B (zh) 一种基于免疫组化生成训练数据的方法和存储设备
Taylor Immunohistochemistry in surgical pathology: principles and practice
CN109946139A (zh) 一种石蜡切片免疫组化套染pas试剂盒及其染色方法和应用
CN105510600A (zh) 晚期乳腺癌患者外周血循环肿瘤细胞er基因的检测方法
Refaiy et al. Semiquantitative smoothelin expression in detection of muscle invasion in transurethral resection and cystectomy specimens in cases of urinary bladder carcinoma
Tolkach et al. Three‐dimensional reconstruction of prostate cancer architecture with serial immunohistochemical sections: hallmarks of tumour growth, tumour compartmentalisation, and implications for grading and heterogeneity
CN106153922A (zh) 一种结肠癌预后预测标志物及其检测方法
Layfield et al. Interobserver reproducibility for HER2/neu immunohistochemistry: a comparison of reproducibility for the HercepTest™ and the 4B5 antibody clone
Tripathy et al. Immunocytochemistry using liquid-based cytology: a tool in hormone receptor analysis of breast cancer
CN112305213A (zh) 富兮双染染色试剂盒及其使用方法
Saiz et al. Infiltrating breast carcinoma smaller than 0.5 centimeters: is lymph node dissection necessary?
WO2024011400A1 (zh) 生成用于计算浸润性乳腺癌细胞Ki-67计数的训练数据的方法、存储设备和试剂盒
Durgapal et al. Assessment of Her‐2/neu status using immunocytochemistry and fluorescence in situ hybridization on fine‐needle aspiration cytology smears: Experience from a tertiary care centre in India
Bauman et al. Quantitation of protein expression and co-localization using multiplexed immuno-histochemical staining and multispectral imaging
CN113687085A (zh) 一种评估实体肿瘤Claudin18.2蛋白表达的方法和应用
CN103471903A (zh) 切片用固定液的配置方法及其应用
Lin et al. Standardization of diagnostic immunohistochemistry
Denice Smith et al. A retrospective review of UroVysion fish interpretations over 8.6 years: A major shift in the patient test population
CN104792988A (zh) 一种用于宫颈上皮内瘤变分级辅助诊断的双染试剂盒及其应用
Wilson et al. Cell blocks in urine cytopathology: do they add value to the diagnosis? A pilot study
JP6468961B2 (ja) ２重染色キット
Erdogan‐Durmus et al. Diagnostic significance of GATA 3, TTF‐1, PTH, chromogranin expressions in parathyroid fine needle aspirations via immuonocytochemical method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22950531

Country of ref document: EP

Kind code of ref document: A1