Pathological digital slice generation method and system based on 3D scanning combined with regional imaging

Abstract

The present invention discloses a method and system for generating pathological digital slices by 3D scanning combined with regional imaging. The method and system obtain 3D data of tissue slices by laser 3D scanning of pathological slices, analyze the effective area of ​​the slices, plan the shortest scanning path and the scanning focal length of each path point, use a microscopic scanning platform to scan the effective area and save it as a dedicated file. When viewing the digital slices, the images of the effective areas are combined into complete digital slices according to their locations, and the blank areas are filled with transparent layers. The method and system disclosed by the present invention have the advantages of high scanning efficiency, clear images, and small file storage capacity.

Description

Technical Field

[0001] The invention belongs to the field of medical pathology, and in particular relates to digitization of pathological sections. Background Art

[0002] Pathological sections are made by dehydrating and embedding tissues, cutting them into thin slices of uniform thickness, staining them, and sealing them on the surface of glass slides. They are often used for microscopic observation of tissue structure and cell morphology to clarify the type of disease. Pathological sections are physical slides that can be digitized by scanning with a high-power microscope. Pathological digital sections are easy to store and exchange data, and have been widely used in scenarios such as remote pathology consultation, digital archiving of pathology sections, and artificial intelligence recognition.

[0003] At present, the digitalization of pathological sections usually uses a 20X or 40X microscope with a CMOS or CCD image sensor to scan line by line (line scanning method) or take pictures piece by piece (surface scanning method), and obtains the whole-field digital section of the pathology (Whole Slide Image, WSI) by splicing the image obtained by scanning or taking pictures. Before starting the scan, the camera configured in the pathology digital section scanner will take a panoramic photo of the pathology section, determine the scanning area based on the contrast between the target and the background in the panoramic photo, and calculate the focus and focal length of the microscope scan. However, there are the following technical defects in using this method to obtain the scanning area and focal length:

[0004] 1. Redundancy in scanning range: The camera takes a picture to obtain the scanning area by calculating the contrast difference between the foreground and the background, which makes it easy to identify non-tissue slices such as stains and marker marks on the slide as areas that need to be scanned. Since the microscope scanning area is fixed to a rectangle, the blank area between multiple tissue slices will also be counted in the scanning range. The scanning movement mode of the microscope objective lens is fixed to a zigzag shape. In order to scan all identified areas, the scanning range will use the outermost edge of all areas as the scanning boundary, which makes the scanning range huge, but the actual useful tissue slice area only occupies a very small part of the entire scanned image. Under the premise that the scanning movement speed of the objective lens is the same, scanning too many blank parts or invalid parts will greatly increase the workload and scanning time, and the scanning efficiency is low;

[0005] 2. Partial image blur: When there are multiple tissue slices encapsulated on the glass slide or a single tissue slice is large in area, the camera can only focus on the main part of it when taking pictures to obtain the focus and focal length. If the thickness of the tissue slice is uneven or the encapsulation is tilted, the area above or below the focus position cannot be focused, and the microscope objective will be out of focus when scanning, resulting in partial image blur;

[0006] 3. Large data storage volume: Since the scan area contains blank areas between tissue slices, the WSI file size often ranges from 100MB to 2GB, which puts a lot of pressure on file transmission and archiving storage;

[0007] Therefore, using a camera to obtain the scanning area and focal length will result in scanning too many invalid slice areas, prolonging the scanning time and increasing the file storage volume. Summary of the invention

[0008] In order to overcome the shortcomings of the prior art, the present invention provides a fast, clear, low-storage 3D scanning combined with regional imaging pathology digital slice generation method and system. The specific scheme is as follows:

[0009] A method for preparing pathological digital slices by 3D scanning combined with regional imaging comprises the following steps:

[0010] S1. Collect 3D data of the tissue slice on the pathological section: Use a micro laser sensor to vertically scan the pathological section, and record the 3D data of the tissue slice on the pathological section along the x, y, and z axes and the coordinates of the position of the center point of the tissue slice;

[0011] S2, scanning path and focus distance planning: according to step S1, the 3D data of the tissue slice on the x, y, and z axes and the position coordinates of the center point of the tissue slice are collected, the x, y axis position and the z axis focus distance corresponding to each minimum resolvable area of ​​the tissue slice are calculated, and the digital slice image slice scanning path and the focus distance of each path point are planned;

[0012] S3, image slice scanning: using the microscope scanning platform to scan the digital slice image slices according to the scanning path and focus distance planned in step S2;

[0013] S4, digital slice file saving: splicing digital slice image slices, generating digital slice local images, and saving the scanning results as a special format file;

[0014] S5. Digital slice image construction: A complete pathology digital slice image is generated based on the x and y axis coordinate positions of the center point of each local image in the pathology digital slice, and the blank parts between the local images are filled with transparent layers to the size of the complete digital slice.

[0015] Furthermore, the working type of the micro laser sensor in step S1 is diffuse reflection line scanning, the reference scanning distance is 50 mm, the z-axis scanning height is 2.6 mm, the x-axis scanning width is 18 mm, the laser wavelength is 405 nm, the laser power is 4.8 mW, the z-axis scanning accuracy is 0.1 μm, the x-axis scanning accuracy is 1 μm, the contour data interval is 20 μm, and the number of contour data is 1000 points.

[0016] Furthermore, in step S2, the planning of the digital slice image slice scanning path adopts Bellman-Ford algorithm to calculate the shortest path between scanning multiple digital slice image slices.

[0017] Furthermore, in step S3, the microscope scanning platform has a z-axis movement accuracy of 0.1 μm, an x-axis and a y-axis movement accuracy of 1 μm, and is equipped with a 20X or 40X objective lens, a scanning resolution of the 20X objective lens of 0.5 μm / pixel or a scanning resolution of the 40X objective lens of 0.25 μm / pixel.

[0018] Furthermore, the scanning information in step S4 is the x-axis and y-axis coordinate values ​​of all digital slice partial images and the center point of each digital slice partial image relative to the location of the complete pathological digital slice, the size value of the complete pathological digital slice, and the scanning resolution value.

[0019] A pathology digital slice generation system combining 3D scanning with regional imaging, comprising:

[0020] At least one 3D scanning module, including

[0021] Micro laser sensor, working state is diffuse reflection line scanning mode, reference scanning distance is 50mm, z-axis scanning height is 2.6mm, x-axis scanning width is 18mm, laser wavelength is 405nm, laser power is 4.8mW, z-axis scanning accuracy is 0.1μm, x-axis scanning accuracy is 1μm, profile data interval is 20μm, and the number of profile data is 1000 points;

[0022] At least one microscopic image acquisition module, comprising

[0023] The microscope scanning platform has a z-axis movement accuracy of 0.1μm, and an x- and y-axis movement accuracy of 1μm. It is equipped with a 20X or 40X objective lens, and the scanning resolution of the 20X objective lens is 0.5μm / pixel or the scanning resolution of the 40X objective lens is 0.25μm / pixel;

[0024] The scanning path planning module calculates the x-axis and y-axis positions and the z-axis focus distance of each minimum resolvable area of ​​the tissue slice based on the collected x, y, and z-axis 3D data of the tissue slice and the coordinates of the center point of the tissue slice, and plans the digital slice image segmentation scanning path and the focus distance of each path point;

[0025] A file storage module, writing all digital slice partial images and the x-axis and y-axis coordinate values ​​of each digital slice partial image relative to the position of the complete pathology digital slice and the size value and scanning resolution value of the complete pathology digital slice into a special format file;

[0026] The image processing module stitches the image slices to generate a digital slice partial image, generates a complete pathology digital slice image according to the x and y axis positions of the digital slice partial images, and fills the unscanned blank parts between adjacent digital slice partial images with a transparent layer to the size of the complete pathology digital slice image.

[0027] The method and system for generating pathological digital slices by 3D scanning combined with regional imaging of the present invention have the following advantages:

[0028] 1. Improve the efficiency of digital slice scanning: Obtain the x, y, and z axis data of tissue slices through laser 3D scanning, obtain the contour information based on the x and y axis data, and obtain the thickness information based on the z axis data to determine whether the scanning target is a tissue slice. 3D scanning can obtain the boundaries of the tissue slice scanning area, its position on the slice, and the focal length, effectively excluding invalid areas such as short fibers with low z-axis thickness values, marker marks, paraffin stains, etc., skip the blank areas between multiple tissue slices through scanning path planning, and only scan the tissue slice area, reducing the invalid scanning area. Under the premise of the same microscopic scanning speed, the scanning time is shorter;

[0029] 2. Improve the clarity of digital slice images: By analyzing the z-axis data of tissue slices obtained by 3D scanning, the focus of each moving path point during microscope scanning is planned so that the scanning plane is always focused on the tissue slice plane, ensuring that the scanned image has high clarity;

[0030] 3. Reduce the capacity of digital slice files: Since the blank areas between tissue slices are skipped during scanning, only the local digital slice images of the tissue slice area are scanned, and then saved in a dedicated file together with the position, complete digital slice size value, and scanning resolution value. When the user uses it, the dedicated software combines the image according to the position of the local digital slice image and the complete slice size value, adds a transparent layer to form a complete digital slice for browsing, omitting the invalid background data in the image, so that the pathology digital slice file occupies less storage space.

[0031] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The following embodiments are only for illustrating the technical ideas of the present invention, and the protection scope of the present invention cannot be limited thereto. Any changes made on the basis of the technical scheme according to the technical ideas proposed by the present invention fall within the protection scope of the present invention. The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic flow chart of the method of the present invention.

[0033] Figure 2 It is a schematic diagram of the system structure of the present invention.

[0034] Figure 3 It is a scanning schematic diagram of the micro laser sensor of the present invention.

[0035] Figure 4 This is a diagram of the scanning results of the micro laser sensor of the present invention.

[0036] Figure 5 It is the scanning path planning diagram of the present invention.

[0037] Figure 6 It is a schematic diagram of the storage content of the special file of the present invention.

[0038] Figure 7 This is a digital slice scanning result diagram of the present invention.

[0039] Figure 8 This is the digital slice result image scanned by traditional method. DETAILED DESCRIPTION

[0040] Taking a pathological slice containing two tissue slices as an example object, the method and system are used to generate a pathological digital slice. The specific implementation method is as follows:

[0041] like Figure 1As shown, the method for generating pathological digital slices by 3D scanning combined with regional imaging of the present invention includes: 3D data acquisition of tissue slices on pathological slices, scanning path and focus distance planning, image slice scanning, digital slice file storage, and digital slice image construction, a total of 5 steps.

[0042] like Figure 2 As shown, the pathology digital slice generation system of 3D scanning combined with regional imaging of the present invention includes: a 3D scanning module 201, a scanning path planning module 202, a microscopic image acquisition module 203, a file storage module 204, and an image processing module 205, a total of 5 modules.

[0043] Specifically, 3D data collection is performed on the tissue slice on the pathological slice: the pathological slice is vertically scanned by the 3D scanning module 201, and the scanning method is as follows: Figure 3 shown.

[0044] It should be noted that the working type of the micro laser sensor of the 3D scanning module 201 is diffuse reflection line scanning, the reference scanning distance is 50mm, the z-axis scanning height is 2.6mm, the x-axis scanning width is 18mm, the laser wavelength is 405nm, the laser power is 4.8mW, the z-axis scanning accuracy is 0.1μm, the x-axis scanning accuracy is 1μm, the contour data interval is 20μm, and the number of contour data is 1000 points.

[0045] The 3D scanning results are as follows:

[0046] The outer contour size of the first tissue slice is 2.21mm×3.45mm, and the average thickness is 6.2μm; the outer contour size of the second tissue slice is 2.44mm×3.52mm, and the average thickness is 4.4μm. The size of the pathological digital slice scanning area is 11.66mm×5.14mm, the position of the center point of the first tissue slice is x=2.34mm, y=2.00mm, and the position of the center point of the second tissue slice is x=10.10mm, y=3.36mm. The contour map is drawn based on the outer contour size, location and thickness data of the tissue slice as shown in the figure. Figure 4 As shown. High-precision 3D laser scanning can obtain 3D information of tissue slices, and analyze the outline size of tissue slices, relative position information, and focus distances of different points.

[0047] Scanning path and focus distance planning: The scanning path planning module 202 calculates the x, y axis position and z axis focus distance of each minimum resolvable area of ​​the tissue slice according to the 3D data of the x, y, and z axes of the first and second tissue slices and the position coordinates of the center point of the tissue slice, and plans the digital slice image slice scanning path and the focus distance of each path point. The scanning path planning module 202 uses the Bellman-Ford algorithm to calculate the shortest path between multiple scanning slices, so that the microscopic image acquisition module 203 has the shortest moving distance when scanning. Through path planning, the microscopic image acquisition module can scan the entire tissue slice with the optimal path, avoid the blank parts between multiple tissue slices, reduce the invalid scanning area, and improve the scanning efficiency. Calculating the tissue thickness of each path point can make the scanning focus of the microscopic image acquisition module 203 always on the surface of the tissue slice, and the acquired image slices will not become blurred due to defocusing.

[0048] like Figure 5 As shown, the scanning plane path planning result 501 is used to guide the movement of the microscopic image acquisition module 203 in the horizontal direction, and the scanning vertical path planning result 502 is used to guide the movement of the microscopic image acquisition module 203 in the vertical direction.

[0049] Image slice scanning: The microscopic image acquisition module 203 provides the path and scan slice image slices according to the scanning plane path planning result 501 and the scanning vertical path planning result 502 planned by the scanning path planning module 202. The image slices are the original linear image data obtained by the microscopic image acquisition module 203 in a line scanning manner or the original block image data obtained by the microscopic image acquisition module 203 in a surface scanning manner.

[0050] It should be noted that the microscope scanning platform has a z-axis movement accuracy of 0.1 μm, and an x- and y-axis movement accuracy of 1 μm. It is equipped with 20X and 40X objective lenses. The scanning resolution of the 20X objective lens is 0.5 μm / pixel, and the scanning resolution of the 40X objective lens is 0.25 μm / pixel. This embodiment uses a 20X objective lens for scanning.

[0051] Digital slice file saving: The image scanning accuracy is 2pixel / μm, the size of the complete digital slice is 23312×10288 pixels, the center point of the first tissue slice is located at [4672,4000] pixels, and the center point of the second tissue slice is located at [20192, 6720] pixels. The digital slice image pixel size and the actual physical slice size can be converted by the formula: image pixel value = image scanning accuracy × physical size value. Figure 6As shown, the first tissue slice image slice and the second tissue slice image slice are spliced ​​to generate a digital slice partial image 601, and the first tissue slice and the second tissue slice digital slice partial image 601, the position of the image center point 602, the complete digital slice size 603 and the scanning resolution 604 are saved as a special format file by the file storage module 204. Compared with the traditional method of saving the entire digital slice including the background, the method of the present invention reduces the storage content and saves storage space.

[0052] Digital slice image construction: Figure 7 As shown, the image processing module 205 generates a complete pathology digital slice image based on the local images and positions of the first tissue slice and the second tissue slice, and fills the blank parts between the local images of the digital slice with transparent layers to the pathology digital slice size of 23312×10288 pixels to form a complete pathology digital slice. The user can also add a white solid color background or a glass slide photo as a virtual background under the complete digital slice with a transparent layer to suit personal reading preferences.

[0053] In order to further compare the scanning results, file storage capacity and scanning speed of the method of the present invention with those of the existing methods, the KF-PRO-005 fully automatic pathology digital slide scanner was used to scan the same slide under the same experimental conditions of 20X magnification and the same microscope movement speed as the method of the present invention. The digital slides obtained by scanning are shown in FIG. Figure 8 shown.

[0054] The KF-PRO-005 fully automatic pathology digital slice scanner uses a snapshot method to obtain the scanning target area. The stains under the tissue slice are also identified as scanning objects, so the scanning range includes more blank and invalid areas. The scanning focus is only focused on the first tissue slice. The second tissue slice is thinner, and defocusing occurs during scanning, so the image of the second tissue slice is blurred. The KF-PRO-005 fully automatic pathology digital slice scanner takes 45 seconds to collect digital slices, and the kfb file capacity for storing pathology digital slices is 131MB; the method of the present invention takes a total of 14 seconds to collect digital slices, of which laser 3D scanning takes 4 seconds, microscope scanning takes 10 seconds, and the lr file capacity for storing pathology digital slices is 81MB. In comparison, the scanning time of the method of the present invention is shortened by 68.9%, saving 38.2% of storage space. This is because under the conditions of the same scanning resolution and scanning speed, the method of the present invention does not include the blank areas and stain areas between tissue slices into the scanning area, saving scanning time and storage space; the focal length of each scanning movement point is obtained through 3D scanning to ensure that the scanning focus is always on the tissue slice, so that large-range scanning images have good clarity.

Claims

1. A method for producing pathological digital slices by 3D scanning combined with regional imaging, comprising the following steps: S1. Collect 3D data of tissue slices on pathological sections: Use a micro laser sensor to vertically scan the pathological sections, and record the 3D data of the tissue slices on the pathological sections along the x, y, and z axes and the coordinates of the position of the center point of the tissue slices; S2, scanning path and focus distance planning: according to step S1, the 3D data of the tissue slice x, y, z axis and the position coordinates of the center point of the tissue slice are collected, the x, y axis position and z axis focus distance corresponding to each minimum resolvable area of ​​the tissue slice are calculated, and the digital slice image segmentation scanning path and the focus distance of each path point are planned; S3, image slice scanning: using the microscope scanning platform to scan the digital slice image slices according to the scanning path and focus distance planned in step S2; S4, digital slice file saving: splicing digital slice image slices, generating digital slice local images, and saving the scanning results as a special format file; S5, digital slice image construction: generate a complete pathology digital slice image according to the x and y axis coordinate positions of the center point of each local image in the pathology digital slice, and fill the blank parts between the local images with transparent layers to the size of the complete digital slice; The x, y, and z-axis data of tissue slices are acquired by laser 3D scanning, and the contour information is acquired based on the x and y-axis data, and the thickness information is acquired based on the z-axis data, so as to determine whether the scanning target is a tissue slice; by analyzing the z-axis data of tissue slices acquired by 3D scanning, the focal length of each moving path point during microscope scanning is planned so that the scanning plane is always focused on the tissue slice plane, ensuring that the scanned image has high clarity; only the local image of the digital slice in the tissue slice area is scanned, and then saved in a dedicated file together with the position, complete digital slice size value, and scanning resolution value.

2. The method for preparing pathological digital slices by 3D scanning combined with regional imaging according to claim 1 is characterized in that : In step S1, the working type of the micro laser sensor is diffuse reflection line scanning, the reference scanning distance is 50 mm, the z-axis scanning height is 2.6 mm, the x-axis scanning width is 18 mm, the laser wavelength is 405 nm, the laser power is 4.8 mW, the z-axis scanning accuracy is 0.1 μm, the x-axis scanning accuracy is 1 μm, the contour data interval is 20 μm, and the number of contour data points is 1000.

3. The method for preparing pathological digital slices by 3D scanning combined with regional imaging according to claim 1 is characterized in that In step S2, the digital slice image slice scanning path is planned by using the Bellman-Ford algorithm to calculate the shortest path between scanning multiple digital slice image slices.

4. The method for preparing pathological digital slices by 3D scanning combined with regional imaging according to claim 1 is characterized in that In step S3, the microscope scanning platform has a z-axis movement accuracy of 0.1 μm, an x-axis and a y-axis movement accuracy of 1 μm, and is equipped with a 20X or 40X objective lens, a 20X objective lens scanning resolution of 0.5 μm / pixel, or a 40X objective lens scanning resolution of 0.25 μm / pixel.

5. The method for producing pathological digital slices by 3D scanning combined with regional imaging according to claim 1 is characterized in that The scanning result in step S4 is the x-axis and y-axis coordinate values ​​of all digital slice partial images and the center point of each digital slice partial image relative to the location of the complete pathological digital slice, the size value of the complete pathological digital slice, and the scanning resolution value.

6. A 3D scanning combined with regional imaging pathology digital slice generation system, which implements the 3D scanning combined with regional imaging pathology digital slice production method of claim 1, comprising: At least one 3D scanning module, including Micro laser sensor, working state is diffuse reflection line scanning mode, reference scanning distance is 50mm, z-axis scanning height is 2.6mm, x-axis scanning width is 18mm, laser wavelength is 405nm, laser power is 4.8mW, z-axis scanning accuracy is 0.1μm, x-axis scanning accuracy is 1μm, contour data interval is 20μm, and round data quantity is 1000 points; At least one microscopic image acquisition module, comprising The microscope scanning platform has a z-axis movement accuracy of 0.1μm, and an x- and y-axis movement accuracy of 1μm. It is equipped with a 20X or 40X objective lens, and the scanning resolution of the 20X objective lens is 0.5μm / pixel or the scanning resolution of the 40X objective lens is 0.25μm / pixel; The scanning path planning module calculates the x-, y-axis position and z-axis focus distance of each minimum resolvable area of ​​the tissue slice based on the collected x, y, and z-axis 3D data of the tissue slice and the coordinates of the center point of the tissue slice, and plans the digital slice image segmentation scanning path and the focus distance of each path point; A file storage module, writing all digital slice partial images and the x-axis and y-axis coordinate values ​​of each digital slice partial image relative to the position of the complete pathology digital slice and the size value and scanning resolution value of the complete pathology digital slice into a special format file; The image processing module generates a digital slice partial image by splicing the image slices, generates a complete pathology digital slice image according to the x and y axis positions of the digital slice partial images, and fills the unscanned blank parts between adjacent digital slice partial images with a transparent layer to the size of the complete pathology digital slice image.

Images