Cervical cancer TCT slide microscope automatic scanning and AI diagnosis system and method

Abstract

The present invention discloses a TCT slide microscope automatic scanning and AI diagnosis system and method for cervical cancer. The system comprises: a TCT slide microscope automatic scanner and an AI cervical cancer cell diagnosis classification and identification system. The method controls the slide sample on the automatic stage of the microscope automatic scanner to move one objective lens window distance each time, and collects and diagnoses the window scanned image. The entire slide is scanned and diagnosed by continuously and uninterruptedly, one objective lens window at a time.

Description

Technical Field

[0001] The present invention relates to the fields of medical treatment and artificial intelligence, and in particular to a system and method for automatic scanning and AI diagnosis of cervical cancer TCT slide microscope. Background Art

[0002] TCT examination uses a liquid-based thin-layer cell detection system to detect cervical cells and perform cytological classification diagnosis. It is currently the most advanced cervical cancer cell cytological examination technology in the world. The current difficulty in providing large-scale TCT testing for women is that the workload of manual cell classification testing is too large, and there is a serious shortage of pathologists, which leads to insufficient testing capabilities of existing specialized hospitals and third-party institutions, too slow testing speed, and the need to improve accuracy.

[0003] An advanced solution is to use an automatic pathology section scanner to scan the entire TCT slide, and then use the cervical cancer screening automatic diagnosis and classification software for diagnosis and classification. The problem is that scanning the entire TCT slide will generate a file of about 10GB of big data, which makes data processing and storage very difficult, and also leads to very high computing and storage costs.

[0004] To solve the above difficulties, the present invention provides a cervical cancer TCT slide microscope automatic scanning and AI diagnosis system and method, which scans each image under the objective window of the microscope TCT slide automatic scanning system separately and completes the AI ​​cervical cancer cell diagnosis classification and identification. Thus, low-cost, fast, high-accuracy and high-efficiency large-scale screening of cervical cancer is achieved. Summary of the invention

[0005] The present invention discloses a system and method for automatic scanning and AI diagnosis of cervical cancer TCT slide microscope. The system comprises: a TCT slide microscope automatic scanner and an AI cervical cancer cell diagnosis classification and identification system. The method controls the slide sample on the automatic stage of the microscope automatic scanner to move one objective lens window distance each time, and collects and diagnoses the window scanned image. The entire slide is completed by continuously and uninterruptedly scanning and diagnosing each objective lens window one by one. The AI ​​cervical cancer cell diagnosis classification and identification system is used to perform AI cervical cancer cell diagnosis classification and identification on the image scanned under each window.

[0006] The TCT slide microscope automatic scanning and AI diagnosis method for cervical cancer includes the following steps:

[0007] Step 1: Control the automatic stage of the scanner to move the slide cell sample to the objective window, take an image of the cell sample, and send it to the AI ​​cervical cancer cell diagnosis classification and identification system to complete the cervical cancer cell diagnosis classification and identification. If abnormal cells are detected, store the image and the diagnosis result; if no abnormal cells are detected, discard the image;

[0008] Step 2: Control the automatic stage of the scanner to move the slide cell sample to a distance of the objective window, and complete the image acquisition and AI cervical cancer cell diagnosis classification and identification under the window. Similarly, if abnormal cells are detected, keep the image and diagnosis results; if no abnormal cells are detected, abandon the image. Continue to move the automatic stage to move the slide cell sample to the next window until the x-axis line scan is completed;

[0009] Step 3: After the slide sample completes the x-axis line scan, the automatic stage continues to move the slide cell sample along the y direction by one window, and then scans back along the x-axis. After completing another line scan, it moves another window distance until all y-axis line scans are completed;

[0010] Step 4: After scanning the entire slide sample, all images containing abnormal cells are stitched together into a complete cell diagnosis and classification map, which is used as an attached figure in the diagnosis result report.

[0011] As a further technical solution of the present invention, in step 1, the AI ​​cervical cancer cell diagnosis, classification and identification system is a cervical cancer cell diagnosis, classification and identification system based on machine learning, which can automatically diagnose, classify and identify the collected cervical cancer cell images, that is, the cervical cancer cell images taken under the microscope window.

[0012] Furthermore, the cervical cancer cells are classified into two categories: normal cells and abnormal cells.

[0013] Furthermore, the abnormal cells are divided into five categories: ASC-US, ASC-CS, ASC-H, LSIL, and HSIL.

[0014] As a further technical solution of the present invention, in step 2, the objective lens window distance in the x-axis direction is , in the y-axis direction , the objective lens window size is .

[0015] Preferably, the scanning starts from the origin of the scanned slide on the automatic stage.

[0016] Furthermore, the automatic stage can be loaded with multiple standard TCT slides, and scans one slide before scanning another.

[0017] Furthermore, the stage increases the viewing window distance along the x-axis direction each time. The stage is moved to ensure that the slide is scanned continuously and without omissions, and each line is scanned and then scanned again until the entire slide is scanned.

[0018] The beneficial effects of the present invention are as follows: a system and method for automatic scanning and AI diagnosis of TCT slide microscope for cervical cancer is provided. The method divides the entire TCT slide into each objective lens window size for scanning window by window, and then uses an AI diagnostic classification and recognition system to diagnose and classify and recognize each scanned image. This solves the problem that ordinary pathology scanners have large amounts of data to process, and that processing and storage are difficult. A low-cost, high-efficiency automatic screening method for cervical cancer is provided, which solves the current pain points of insufficient resources, high costs, and insufficient accuracy for cervical cancer screening. Large-scale promotion of the present invention can save a lot of manpower and material resources, serve the general public, and achieve huge social and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of a cervical cancer TCT slide microscope automatic scanning and AI diagnosis system and method according to a preferred embodiment of the present invention;

[0020] Figure 2 This is the workflow of the TCT slide microscope automatic scanning and AI diagnosis system and method for cervical cancer in the preferred embodiment of the present invention. DETAILED DESCRIPTION

[0021] The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments. These drawings are simplified schematic diagrams, which only illustrate the basic principles, basic structures and basic functions of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0022] A person skilled in the art may be aware of some variations and equivalent substitutions of the present invention, but these variations and equivalent substitutions should not be construed as exceeding the protection scope of the present invention.

[0023] Figure 1 This is a schematic diagram of the automatic scanning and AI diagnosis system and method of TCT slide microscope for cervical cancer according to the preferred embodiment of the present invention. Figure 1 As shown, the system consists of a TCT slide microscope automatic scanner 1 and an AI cervical cancer cell diagnosis, classification and identification system 2. Under the control of 2, 1 moves one microscope objective window at a time to complete window-by-window and line-by-line scanning. When scanning each window point, 2 collects the image under 1 window to perform AI cervical cancer diagnosis and identification.

[0024] Figure 2This is the workflow of the TCT slide microscope automatic scanning and AI diagnosis system and method for cervical cancer in the preferred embodiment of the present invention. Figure 2 As shown, process S01 controls the automatic stage slide cell sample of the scanner to move to the objective window, focuses to make the image clear, captures the image under the window, and sends the image to the AI ​​cervical cancer cell diagnosis classification and identification system for cervical cancer cell diagnosis classification and identification; process S02: if abnormal cells are detected in S01, the image and the diagnosis result are stored; if no abnormal cells are detected, the image is abandoned; process S03: control the slide sample on the automatic stage to move along the x direction by a distance of the objective window; process S04: determine whether the x-line scan is completed. If not finished; enter process S05: control the slide sample on the stage to move to the next window along the x-axis; return to process S01; if S04 determines that the x-line scan is finished; enter process S06: determine whether the y-direction scan is completed; if the scan is not completed; enter process S07: move the slide sample on the stage along the y direction for a window distance, enter the next line scan, and at this time x is scanned in the opposite direction; if the result of S06 is determined to be finished; enter process S09: the scan of the entire slide is completed.

[0025] The above embodiments are descriptions of specific implementation methods of the present invention rather than limitations of the present invention. Technical personnel in the relevant technical field can make various changes and modifications to obtain corresponding equivalent technical solutions without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions should be included in the patent protection scope of the present invention.

Claims

1. A diagnostic method applied to the automatic scanning of TCT slide microscope and AI diagnosis system for cervical cancer, characterized in that: The cervical cancer TCT slide microscope automatic scanning and AI diagnosis system includes a TCT slide microscope automatic scanner and an AI cervical cancer cell diagnosis classification and recognition system; The diagnostic method comprises the following steps: Step 1: Control the automatic stage of the scanner to move the slide cell sample to the objective window, take an image of the cell sample, and send it to the AI ​​cervical cancer cell diagnosis, classification and identification system to complete the diagnosis, classification and identification of cervical cancer cells; if abnormal cells are detected in the objective window, store the image and the diagnosis result; if no abnormal cells are detected, discard the image; Step 2: Control the automatic stage of the scanner to move the glass slide cell sample along the x-axis direction by the distance of an objective window, complete the image acquisition under the objective window and the AI ​​cervical cancer cell diagnosis classification and identification, if the objective window detects abnormal cells, store the image and the diagnosis result, if no abnormal cells are detected, abandon the image; continue to move the automatic stage to move the glass slide cell sample to the next objective window, complete the image acquisition under the objective window and the AI ​​cervical cancer cell diagnosis classification and identification, if the objective window detects abnormal cells, store the image and the diagnosis result, if no abnormal cells are detected, abandon the image, until the x-axis direction line scan is completed; In step 2, the objective lens window distance in the x-axis direction is , in the y-axis direction , the objective lens window size is ; Step 3: After the glass slide cell sample has completed scanning in the x-axis direction, the automatic stage is continuously moved to move the glass slide cell sample along the y-axis by one objective lens window, and then the glass slide cell sample is retraced along the x-axis direction. After completing another line of scanning, the glass slide cell sample is moved along the y-axis by one window distance until all lines of scanning along the y-axis are completed. If abnormal cells are detected in the objective lens window, the image and the diagnosis result are stored. If no abnormal cells are detected, the image is discarded. Step 4: After the entire slide cell sample is scanned, all images containing abnormal cells are spliced ​​into a complete cell diagnosis and classification map, which is used as an accompanying figure in the diagnosis result report.

2. The diagnostic method for cervical cancer TCT slide microscope automatic scanning and AI diagnostic system according to claim 1, characterized in that: In the step 1, only one image in the objective lens window under the microscope objective lens is scanned, collected and diagnosed each time.

3. The diagnostic method for cervical cancer TCT slide microscope automatic scanning and AI diagnostic system according to claim 1, characterized in that: The AI ​​cervical cancer cell diagnosis, classification and identification system is a cervical cancer cell diagnosis, classification and identification system based on machine learning, which automatically diagnoses, classifies and identifies cervical cancer cell images.

4. The diagnostic method for cervical cancer TCT slide microscope automatic scanning and AI diagnostic system according to claim 1, characterized in that: In step 2, the automatic stage of the scanner moves the slide cell sample along the x-axis direction one objective lens window at a time, and moves continuously and uninterruptedly to avoid missing scans.

5. The diagnostic method for cervical cancer TCT slide microscope automatic scanning and AI diagnostic system according to claim 1, characterized in that: In step 4, the diagnosis result report includes the diagnosis result and a drawing, wherein the drawing is a cell diagnosis and classification diagram spliced ​​together from all the images containing the abnormal cell diagnosis results.

Images