Tissue slice imaging method and imaging system

Abstract

The invention provides a tissue slice imaging method and an imaging system, and relates to the technical field of biomedical detection. The tissue slice imaging method specifically comprises the following steps: step 1, slice acquisition; 2, performing two-dimensional imaging; 3, performing two-dimensional splicing and combination; and 4, splicing and verifying the three-dimensional model. A tissue slice imaging system includes an intelligent imaging system. According to the method, an attached structure block is used for carrying out attached addition on a slice structure and then carrying out outer contour scanning recording, feature matching is carried out based on feature association of a sphere block fragment and an SIFT registration algorithm, and geometric correction is carried out on image matching by using a PROSAC algorithm, so that the precision and accuracy of image recombination synthesis are effectively improved; and two-dimensional imaging and three-dimensional visual imaging technologies are realized, and final verification is carried out, so that the reliability of three-dimensional imaging is improved.

Description

Technical Field

[0001] The invention relates to the technical field of biomedical detection, and in particular to a tissue slice imaging method and an imaging system. Background Art

[0002] The mainstream method of existing pathological tissue imaging is microscopic imaging technology based on visible light. Although the spatial resolution of optical imaging methods is very high (submicron), the penetration of visible light is poor, and only very thin pathological sections (greater than tens of microns) can be directly imaged. As the thickness of the pathological sections increases, the images between cells may overlap, and the global three-dimensional structural information of the entire pathological tissue cannot be obtained. In order to reduce the overlap of cell images, the pathological sections are usually only a few microns (smaller than the cell size). Obviously, a moderately sized pathological tissue can produce a huge number of pathological sections.

[0003] The prior art (CN116263414A) is a pathological sample imaging system, imaging method and storage medium. The reconstructed pathological image can quickly and accurately locate the lesion area of ​​interest in the embedded sample, assist in the accurate slicing of the tissue corresponding to the area of ​​interest in the later stage, and perform conventional optical microscopic imaging, thereby improving the efficiency of pathological slice preparation and imaging observation; the prior art (CN112683807A) is a tissue slice imaging method and imaging system, which uses dual-band excitation photoacoustic microscopic imaging technology to obtain label-free high-resolution and high-contrast tissue slice imaging, which can be widely used in the detection field; although the prior art can realize the tissue slice imaging function, it can only complete the single-shot single-slice imaging function, lacks the three-dimensional visualization microscopic imaging effect after the entire tissue block is sliced, and does not have a visualization interactive interface, lacks interactive functions and comparison functions, and cannot realize the auxiliary diagnosis and judgment function. Summary of the invention

[0004] (I) Technical solution

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a tissue slice imaging method, specifically comprising the following steps:

[0006] Step 1: Slice acquisition:

[0007] Select a tissue block that needs to be sliced ​​and imaged, and perform HE preparation on the tissue block, wherein before performing the transparent wax dipping step, first add a spherical block with a height greater than the thickness of the tissue block and with the upper and lower ends removed to the outer wall of the tissue block after the dehydration process so that the spherical block fits the outer wall of the tissue block, and the upper and lower ends of the spherical block are both higher than the tissue block, and the outer contour of the added structure is scanned and recorded; then complete the embedding, slicing and sealing steps;

[0008] Step 2: 2D imaging:

[0009] Placing multiple tissue coverslips of embedded sections on multiple slides at equal intervals in sequence and placing multiple slides at equal intervals on an electric stage, using the electric stage to complete the clockwise rotation of the multiple slides, and having a CCD camera at the lower end to continuously take pictures, and having a first objective lens and an optical microscope at the upper end to directly observe;

[0010] When taking pictures and observing the limited position of slices, the automatic focusing system on one side emits infrared light, which is refracted by the dichroic mirror and irradiated on the tissue slice through the focusing objective lens to achieve auxiliary positioning effects for fluorescence positioning and objective lens adjustment;

[0011] Step 3: 2D splicing combination:

[0012] Based on the spherical block added in step 1, the first tissue slice and the terminal tissue slice of the multiple tissue blocks that have completed the slicing both contain a convex block, and the tissue slices are feature-associated with the fragments of the spherical block, feature matching is performed based on the SIFT registration algorithm, and image recombinant synthesis is performed after image matching. According to the correspondence between the features of each tissue slice, a geometric transformation model is selected, and the parameters of the geometric transformation model are optimized using two similarity functions to perform image registration;

[0013] Step 4: 3D model splicing verification:

[0014] Based on the image registration information in step three, the two-dimensional data images of multiple tissue slices are superimposed and spliced ​​in sequence, and three-dimensional visualization is performed using surface drawing technology after local optimization using linear interpolation; the coordinate alignment and verification of the spliced ​​and reorganized three-dimensional visualization data image and the three-dimensional structure recorded by the outer contour scan reserved in step one are performed, and finally a decomposable three-dimensional visualization image based on a combination of multiple blocks is generated.

[0015] Preferably, the recording method of the outer contour scanning record in step 1 includes:

[0016] Based on the active contour model method, the contour scanning technology is used to obtain the increased structural morphological outer contour, an image definition template is generated based on the structural morphological outer contour, and the model segmentation contour is defined according to the thickness of each slice in the slicing process to complete the data recording.

[0017] Preferably, the spherical block is made of a colloidal material sphere, which ensures that the surface of the tissue block is flat when it is fitted with the tissue block, and the outer walls of the two are always fitted and connected during wax dipping.

[0018] Preferably, the automatic focusing system can emit infrared light of multiple different wavelengths, and the optical microscope and CCD camera are both connected to a computer.

[0019] Preferably, the clockwise rotation of the electric stage in the step two is controlled based on a miceomanager software control system, in which a plurality of slides are divided into a plurality of areas, and timing and serial numbering of the execution shooting time of each divided area are performed, wherein the timing of each area includes the customization of the time required from the end point of the previous shooting to the initial starting point of the next area after the shooting of the previous area is completed, and the customization of the time from the initial point to the end point of the area.

[0020] Preferably, the multiple tissue slices are arranged end to end with the raised blocks of the first tissue slice and the end tissue slice in step three as dividing points, and the slides of the first tissue slice and the end tissue slice are provided with reserved openings so that the raised blocks pass through the reserved openings to ensure that they fit on the slide.

[0021] Preferably, based on the relative arrangement of the raised blocks of the first tissue slice and the end tissue slice, when the first tissue slice is attached to the slide, the image presented by the first tissue slice is a back image, that is, when performing image matching in step three, the image of the first tissue slice is mirror-flipped before image matching is performed.

[0022] Preferably, a tissue slice imaging system includes an intelligent imaging system, wherein the intelligent imaging system includes:

[0023] A two-dimensional imaging unit, used to realize imaging of a single tissue slice;

[0024] A three-dimensional stitching imaging unit is used to complete three-dimensional visualization imaging by superimposing and stitching multiple two-dimensional images of multiple tissue slices;

[0025] User interaction interface unit, used to complete the interactive functions of terminal visualization, online human-computer or human-human interaction;

[0026] A database comparison unit, used for auxiliary diagnosis reference after comparing and analyzing the slice imaging information with the data in the database;

[0027] Communication unit, completing the communication function between multiple devices;

[0028] Data acquisition unit, which collects information between multiple devices;

[0029] A data processing unit, used for analyzing and processing the data information collected by the data acquisition unit;

[0030] Storage module, used to complete data and system function storage.

[0031] (II) Beneficial effects

[0032] The present invention provides a tissue slice imaging method and imaging system, which have the following beneficial effects:

[0033] 1. The present invention provides a tissue slice imaging method and imaging system, which uses an additional structural block to add an additional slice structure and then scans and records the outer contour, providing a precise reference for subsequent image stitching and verification, and improving the accuracy and integrity of imaging. An electric stage is used to rotate multiple slides clockwise, combined with continuous image capture by a CCD camera and direct observation with an optical microscope, and feature matching is performed based on the feature association of spherical block fragments and the SIFT registration algorithm, effectively improving the precision and accuracy of image recombination and synthesis. At the same time, the two-dimensional data graphs are superimposed and spliced ​​in sequence, and linear interpolation is used for local optimization and surface drawing technology to achieve three-dimensional visualization, and coordinate alignment verification is performed with the three-dimensional structure recorded by the outer contour scan, thereby improving the reliability of three-dimensional imaging.

[0034] 2. The present invention provides a tissue slice imaging method and imaging system. The tissue slice imaging system covers multiple units such as two-dimensional imaging, three-dimensional splicing imaging, user interaction, database comparison, communication, data acquisition and processing, and storage, realizing the full process function from image acquisition to analysis and processing, and providing an integrated solution for tissue slice imaging. The system can support multiple functions, improve the communication efficiency between doctors and patients, and also realize intelligent pathological analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the framework structure of the method process of the present invention;

[0036] Figure 2 This is a schematic diagram of the tissue slice image acquisition framework structure of the present invention;

[0037] Figure 3 Schematic diagram of the framework structure of the intelligent imaging system of the present invention.

[0038] Among them, 1. the first objective lens; 2. the optical microscope; 3. the specimen slide; 4. the electronic stage; 5. the focusing objective lens; 6. the dichroic mirror; 7. the filter; 8. the lens; 9. the CCD camera; 10. the light source transmitter; 11. the computer. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] According to the attached Figure 1 As can be seen from the flowchart on the left, the operation process of traditional FFPE-HE staining is simply as follows: tissue fixation - dehydration and transparency - wax immersion - embedding - sectioning - xylene dewaxing - gradient alcohol elution of xylene - water washing - hematoxylin staining - water washing - differentiation - blueing - eosin staining - dehydration - transparent sealing - observation;

[0041] Specifically, sampling and fixation: select the required tissue block, fix it for 12 hours after sampling (if the sample is obtained by 4% paraformaldehyde perfusion, fix it for 6 to 8 hours), rinse it with running water for 1 hour, and place it in 70% to 80% ethanol for long-term storage;

[0042] Dehydration and embedding: 70% ethanol for 1 hour; 80% ethanol for 0.5 hour; 85% ethanol for 0.5 hour; 90% ethanol for 0.5 hour; 95% ethanol (second pass) I for 0.5 hour, II for 0.5 hour; absolute ethanol (second pass) I for 0.5 hour, II for 0.5 hour; xylene (second pass) I for 0.5 hour, II for 0.5 hour (Note: if the tissue block is too large, the dehydration and transparency time can be appropriately extended); wax immersion (50-52°C) (third pass) I for 60 minutes, II for 60 minutes, III for 60 minutes;

[0043] Slicing: trim, cut, expand, stick, bake (baking slices at 50-60℃) for 1h;

[0044] HE staining: xylene dewaxing three times, 5-10 minutes each; anhydrous ethanol I 5min, II 5min; 95% ethanol I5min, II 5mn; 80% ethanol 5min; 70% ethanol 5min; purified water 3-5min; Harris hematoxylin solution 5min (can be extended appropriately in winter, such as 20min); water washing 1-3min; 0.5% hydrochloric acid alcohol color separation 3-10seconds, observe under the microscope; saturated disodium hydrogen phosphate solution blueing 5min; 70% ethanol 5min; 80% ethanol 5min; eosin solution (95% ethanol solution) 3-30seconds;

[0045] Baked slices, sealed with neutral gum

[0046] Among them, hematoxylin is used to stain the cell nucleus: the slices are stained with Harris hematoxylin for 5-10 minutes, washed with tap water for 1 minute, differentiated with 1% hydrochloric acid alcohol for a few seconds, rinsed with tap water for 1 minute, PBS blued for 5 minutes, and rinsed with running water; (9) Eosin staining of cytoplasm: the slices are stained with eosin staining solution for 1-3 minutes; (10) Dehydration and sealing: the slices are placed in 95% alcohol I for 5 minutes-95% alcohol II for 5 minutes-absolute ethanol I for 5 minutes-absolute ethanol II for 5 minutes-xylene I for 5 minutes-xylene II for 5 minutes in sequence for dehydration and transparency, the slices are taken out of xylene and slightly dried, sealed with neutral gum, and dried; finally, the slices are examined under a microscope and image acquisition and analysis is performed.

[0047] Example:

[0048] like Figure 1-2 As shown, an embodiment of the present invention provides a tissue slice imaging method, which specifically includes the following steps:

[0049] Step 1: Slice acquisition:

[0050] Select a tissue block that needs to be sliced ​​and imaged, and perform HE preparation on the tissue block, wherein before performing the transparent wax dipping step, first add a spherical block with a height greater than the thickness of the tissue block and with the upper and lower ends removed to the outer wall of the tissue block after the dehydration process so that the spherical block fits the outer wall of the tissue block, and the upper and lower ends of the spherical block are both higher than the tissue block, and the outer contour of the added structure is scanned and recorded; then complete the embedding, slicing and sealing steps;

[0051] According to the attached Figure 1 From the process structure diagram on the right, we can see that:

[0052] Sampling and fixation: Select the required tissue block, and prepare a spherical block with a height greater than the thickness of the tissue block and the upper and lower ends removed, and place the spherical block tightly against the outer wall of the tissue block, wherein the spherical block is made of a colloid material sphere, which ensures that the surface of the tissue block is flat when it is fitted with the tissue block, and the outer walls of the two are always fitted and connected during wax dipping, and the upper and lower ends of the spherical block both protrude from the upper and lower ends of one side of the tissue block, and the outer contour structure of the tissue block and the spherical block at this moment is scanned and recorded based on the outer contour scanning technology. The outer contour scanning technology can be used for rapid scanning by a profilometer and other equipment, and the scanned data is directly transmitted to the system storage space in the computer for storage and recording;

[0053] The recording method of the outer contour scanning record includes:

[0054] Based on the active contour model method, the contour scanning technology is used to obtain the outer contour of the increased structural morphology, and an image definition template is generated based on the outer contour of the structural morphology. The model segmentation contour is defined according to the thickness of each slice in the slicing process to complete data recording;

[0055] After the tissue block is collected, it is fixed for 12 hours (if the sample is collected by perfusion with 4% paraformaldehyde, it is fixed for 6 to 8 hours), rinsed with running water for 1 hour, and placed in 70% to 80% ethanol for long-term storage.

[0056] Dehydration and embedding: 70% ethanol for 1 hour; 80% ethanol for 0.5 hour; 85% ethanol for 0.5 hour; 90% ethanol for 0.5 hour; 95% ethanol (second pass) I for 0.5 hour, II for 0.5 hour; absolute ethanol (second pass) I for 0.5 hour, II for 0.5 hour; xylene (second pass) I for 0.5 hour, II for 0.5 hour (Note: if the tissue block is too large, the dehydration and transparency time can be appropriately extended); wax immersion (50-52°C) (third pass) I for 60 minutes, II for 60 minutes, III for 60 minutes;

[0057] Slicing: trim, cut, expand, stick, bake (baking slices at 50-60℃) for 1h;

[0058] HE staining: xylene dewaxing for five times, 5-10 minutes for each time; anhydrous ethanol I for 5 minutes, II for 5 minutes; 95% ethanol I for 5 minutes, II for 5 minutes; 80% ethanol for 5 minutes; 70% ethanol for 5 minutes; purified water for 3-5 minutes; Harris hematoxylin solution for 5 minutes (can be extended appropriately in winter, such as 20 minutes); washing with water for 1-3 minutes; 0.5% hydrochloric acid alcohol color separation for 3-10 seconds, observation under microscope; saturated disodium hydrogen phosphate solution for blueing for 5 minutes; 70% ethanol for 5 minutes; 80% ethanol for 5 minutes; eosin solution (95% ethanol solution) for 3-30 seconds;

[0059] Baked slices, sealed with neutral gum

[0060] Among them, hematoxylin is used to stain the cell nucleus: the slices are stained with Harris hematoxylin for 5-10 minutes, washed with tap water for 1 minute, differentiated with 1% hydrochloric acid alcohol for a few seconds, rinsed with tap water for 1 minute, PBS blued for 5 minutes, and rinsed with running water; (9) eosin staining of cytoplasm: the slices are stained with eosin staining solution for 1-3 minutes; (10) dehydration and sealing: the slices are placed in 95% alcohol I for 5 minutes-95% alcohol II for 5 minutes-absolute ethanol I for 5 minutes-absolute ethanol II for 5 minutes-xylene I for 5 minutes-xylene II for 5 minutes in sequence for dehydration and transparency, the slices are taken out of xylene and slightly dried, sealed with neutral gum, and dried;

[0061] Step 2: 2D imaging:

[0062] Placing multiple tissue coverslips of embedded sections on multiple slides at equal intervals in sequence and placing multiple slides at equal intervals on an electric stage, using the electric stage to complete the clockwise rotation of the multiple slides, and having a CCD camera at the lower end to continuously take pictures, and having a first objective lens and an optical microscope at the upper end to directly observe;

[0063] The electric stage is used to automatically and accurately control the shooting position, wherein the clockwise rotation of the electric stage is controlled based on the miceomanager software control system, in which multiple slides are divided into multiple areas, and the timing and sequence number of the execution shooting time of each area are set for each divided area, wherein the timing time of each area includes the customization of the time required from the end point of the previous shooting to the initial starting point of the next area after the completion of the shooting of the previous area, and the customization of the time from the initial point to the end point of the area;

[0064] The user can select or enter the coordinates of the area (such as the upper left corner and the lower right corner) through the interface, and the system stores these areas as a list or array. Each area can have a unique identifier (ID);

[0065] The user can set two time parameters for each area: interval time: the time from the end of shooting in the previous area to the start of shooting in the next area; shooting time: the duration of shooting in the area. These time parameters are stored in a data structure (such as a dictionary or a custom object). By traversing the area list, shooting is performed in sequence according to the time parameters set by the user. The Micromanager API is used to control the camera shooting and implement timing between shooting. After each shooting is completed, the shooting sequence number and time are recorded to ensure that the shooting status of each area can be tracked;

[0066] For example: Area 1: Interval time: 2 seconds, Shooting time: 5 seconds;

[0067] That is, by using precise control shooting at a fixed time and area, the beginning and end fusion stitching order of the front and back pictures can be better completed in the subsequent stitching process, ensuring the integrity, smoothness and authenticity of the stitched image.

[0068] When taking pictures and observing the limited position of slices, the automatic focusing system on one side emits infrared light, which is refracted by a dichroic mirror and then shines through the focusing objective lens onto the tissue slice to achieve auxiliary fluorescence positioning and auxiliary positioning effects for objective lens adjustment. That is, the automatic focusing system emits infrared light, which is refracted by a dichroic mirror onto the sample surface, and the information reflected back is collected to make quick response adjustments, thereby completing the auxiliary analysis of objective lens height adjustment, thereby preventing defocusing caused by large-scale movement; the filter is based on an electric conversion filter box, which automatically switches to capture images produced by lasers of different wavelengths.

[0069] Step 3: 2D splicing combination:

[0070] Based on the spherical block added in step 1, the first tissue slice and the terminal tissue slice of the multiple tissue blocks that have completed the slicing both contain a convex block, and the tissue slices are feature-associated with the fragments of the spherical block, feature matching is performed based on the SIFT registration algorithm, and image recombinant synthesis is performed after image matching. According to the correspondence between the features of each tissue slice, a geometric transformation model is selected, and the parameters of the geometric transformation model are optimized using two similarity functions to perform image registration;

[0071] When cutting the combined body of the tissue block and the spherical block, due to the protruding structure of the spherical block, the first tissue slice and the last tissue slice are both provided with protruding protruding blocks, that is, when the slicing of the entire tissue block is completed, the protruding blocks of the first tissue slice and the last tissue slice are used as the dividing points, and the multiple tissue slices are arranged in sequence head to tail, and the specimen slides of the first tissue slice and the last tissue slice are provided with reserved openings so that the protruding blocks pass through the reserved openings to ensure that they are attached to the specimen slide. Based on the relative arrangement of the protruding blocks of the first tissue slice and the last tissue slice, when the first tissue slice is attached to the specimen slide, the image presented by the first tissue slice is a back image, that is, when performing image matching in step 3, the image of the first tissue slice is mirror-flipped before performing image matching;

[0072] When using feature association, the block after spherical block segmentation is used as the feature point. First, the obvious spherical block features are extracted from the two-dimensional image of each tissue slice to form a feature set; then, the SIFT-based registration algorithm is used to perform feature matching in the feature set corresponding to the two-dimensional image of each tissue slice; when extracting the features of the spherical block, the morphological processing segmentation algorithm is used to extract the spherical block in the tissue slice image, and the boundary of the spherical block is obtained based on contour extraction. For each scanned tissue slice image, the SIFT feature descriptor is used to perform feature matching;

[0073] SIFT feature descriptors are used for feature matching. After feature matching is completed, an affine transformation model is used for registration, wherein the affine transformation model determines the similarity of images using a structural similarity function, and optimizes the model based on a boundary optimization method; that is, firstly, the target image and the source image are loaded, feature points are extracted and matched, the affine transformation matrix is ​​calculated using the matched feature points, and the calculated affine transformation is applied to the source image to generate a registered image; wherein the affine transformation model uses SSIM to optimize the transformation model parameters and then outputs the final registration result using the optimized model;

[0074] When implementing image feature matching, it is necessary to correct and calibrate the matched image, that is, use a PROSAC algorithm (Progressive Sample Consensus) for image matching correction. This algorithm is an improved algorithm based on RANSAC, which mainly solves the efficiency problem of traditional RANSAC when processing data sets. PROSAC first sorts all data points according to the distance error, and the points with higher quality are placed in front. Then, in each iteration, the samples with the best quality in the sorting are selected for model fitting. With each iteration, PROSAC gradually adds more sample points and updates the model, and finally obtains an optimal model; finally, the corrected image is selected through the graph cut method, the stitching seams are automatically selected, and the fusion is achieved according to the multi-bandblending strategy;

[0075] Step 4: 3D model splicing verification:

[0076] Based on the image registration information in step 3, the two-dimensional data images of multiple tissue slices are sequentially superimposed and spliced, and three-dimensional visualization is performed using surface rendering technology after local optimization using linear interpolation; the coordinate alignment and verification of the spliced ​​and reorganized three-dimensional visualization data image and the three-dimensional structure recorded by the reserved outer contour scan in step 1 are performed, and finally a separable three-dimensional visualization imaging image based on a combination of multiple blocks is generated;

[0077] The registered images are stitched together horizontally or vertically into a complete image, and linear interpolation is used for local optimization to improve the stitching quality; that is, the stitching area, especially the seam position that needs to be optimized, is determined, the pixel values ​​on both sides of the seam are extracted from the stitched image, and the seam area is smoothly transitioned through linear interpolation to merge the optimized seam area with the original stitched image; a new image is created whose height is the higher of the two images and whose width is the sum of the widths of the two images minus the width of the overlapping part, and linear interpolation is performed in the overlapping area, and each column of the overlapping area is interpolated by looping to generate new pixel values, and the processed overlapping area is merged into the new image, and the remaining part of the right image is filled in;

[0078] Convert the spliced ​​2D image data into 3D data, interpolate and apply surface rendering technology; use multiple 2D image data from the same tissue block as basic data to generate simulated slice image data, use 3D mimicry technology to generate x and y grid coordinates, create a uniform grid pixel by pixel, assign the intensity value of the image directly to the z-axis to create a 3D effect, and use the 3D drawing library to draw the 3D surface;

[0079] Align the 3D visualization result with the 3D structure of the outer contour to generate a separable 3D visualization image; load and preprocess the reconstructed 3D data and outer contour data to ensure that the two sets of data are in the same coordinate system, apply the ICP algorithm for alignment, set a matching threshold (such as the distance between feature points), and initialize the transformation matrix. Call the registration_icp method to perform alignment, return the transformation matrix, use the transform method to apply the reconstructed data to the obtained transformation matrix to achieve alignment, visualize the aligned reconstructed data and outer contour data in the same coordinate system, and check the alignment effect.

[0080] As attached Figure 3 As shown, a tissue slice imaging system includes an intelligent imaging system, characterized in that the intelligent imaging system includes:

[0081] A two-dimensional imaging unit, used to realize imaging of a single tissue slice;

[0082] A three-dimensional stitching imaging unit is used to complete three-dimensional visualization imaging by superimposing and stitching multiple two-dimensional images of multiple tissue slices;

[0083] The user interaction interface unit is used to complete the interactive functions of terminal visualization, online human-computer or human-human interaction; the user interaction interface unit includes submodules: login / registration interface, real-time video call interface, slice image viewing and annotation interface, chat window and medical record management interface;

[0084] A database comparison unit, used for auxiliary diagnosis reference after comparing and analyzing the slice imaging information with the data in the database;

[0085] The database comparison unit obtains imaging data from the slice imaging system and prepares a database containing existing diagnostic information, such as pathological images, case summaries and known disease markers. It also performs standardization processing on the imaging data, such as scale normalization, denoising, feature extraction, etc., to facilitate subsequent comparison. It extracts key information and features from the slice imaging, which may include color histograms, texture features, morphological features, etc., and compares the extracted features with the feature data in the database. Based on the comparison results, it generates a diagnosis support report, including similar cases, possible diagnostic results, recommended next steps, etc. At the same time, a user interface is created within the system interface, so that doctors can easily input data, view comparison results and related information.

[0086] Communication unit, completing the communication function between multiple devices;

[0087] Data acquisition unit, which collects information between multiple devices;

[0088] A data processing unit is used to analyze and process the data information collected by the data acquisition unit; use machine learning to identify and analyze slice images, including image preprocessing, model training and optimization, and result generation;

[0089] The storage module is used to complete the data and system function storage; wherein the storage module contains multiple submodules: namely, user data table, pathological image data table and chat record table;

[0090] When the system is used, the user name and password are entered through the UI module, and the data is transmitted to the back-end service module for verification. After logging in, the doctor can select the slice to be viewed through the UI module, and the back-end service module requests the relevant images in the storage module. The doctor can start the video call function to communicate with the patient or other doctors in real time and share the slice images. During this period, the intelligent analysis module can be used for image analysis, and the results can be fed back to the UI module in real time.

[0091] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A tissue section imaging method, characterized in that: The specific steps include: Step 1: Slice acquisition: Select a tissue block that needs to be sliced ​​and imaged, and perform HE preparation on the tissue block, wherein before performing the transparent wax dipping step, first add a spherical block with a height greater than the thickness of the tissue block and with the upper and lower ends removed to the outer wall of the tissue block after the dehydration process so that the spherical block fits the outer wall of the tissue block, and the upper and lower ends of the spherical block are both higher than the tissue block, and the outer contour of the added structure is scanned and recorded; then complete the embedding, slicing and sealing steps; Step 2: 2D imaging: Placing multiple tissue coverslips of embedded sections on multiple slides at equal intervals in sequence and placing multiple slides at equal intervals on an electric stage, using the electric stage to complete the clockwise rotation of the multiple slides, and having a CCD camera at the lower end to continuously take pictures, and having a first objective lens and an optical microscope at the upper end to directly observe; When taking pictures and observing the limited position of slices, the automatic focusing system on one side emits infrared light, which is refracted by the dichroic mirror and irradiated on the tissue slice through the focusing objective lens to achieve auxiliary positioning effects for fluorescence positioning and objective lens adjustment; Step 3: 2D splicing combination: Based on the spherical block added in step 1, the first tissue slice and the terminal tissue slice of the multiple tissue blocks that have completed the slicing both contain a convex block, and the tissue slices are feature-associated with the fragments of the spherical block, feature matching is performed based on the SIFT registration algorithm, and image recombinant synthesis is performed after image matching. According to the correspondence between the features of each tissue slice, a geometric transformation model is selected, and the parameters of the geometric transformation model are optimized using two similarity functions to perform image registration; Step 4: 3D model splicing verification: Based on the image registration information in step three, the two-dimensional data images of multiple tissue slices are superimposed and spliced ​​in sequence, and three-dimensional visualization is performed using surface drawing technology after local optimization using linear interpolation; the coordinate alignment and verification of the spliced ​​and reorganized three-dimensional visualization data image and the three-dimensional structure recorded by the outer contour scan reserved in step one are performed, and finally a decomposable three-dimensional visualization image based on a combination of multiple blocks is generated.

2. A tissue slice imaging method according to claim 1, characterized in that: The recording method of the outer contour scanning record in step 1 includes: Based on the active contour model method, the contour scanning technology is used to obtain the increased structural morphological outer contour, an image definition template is generated based on the structural morphological outer contour, and the model segmentation contour is defined according to the thickness of each slice in the slicing process to complete the data recording.

3. A tissue slice imaging method according to claim 1, characterized in that: The spherical block is made of a colloidal material sphere, which ensures that the surface of the tissue block is flat when it is attached to the tissue block, and the outer walls of the two are always attached and connected during wax dipping.

4. A tissue slice imaging method according to claim 1, characterized in that: The automatic focusing system can emit infrared light of multiple different wavelengths, and the optical microscope and the CCD camera are both connected to a computer.

5. A tissue slice imaging method according to claim 1, characterized in that: The clockwise rotation of the electric stage in the step 2 is controlled based on the miceomanager software control system, in which a plurality of slides are divided into a plurality of areas, and the timing and serial numbering of the shooting execution time of each area are performed, wherein the timing time of each area includes the customization of the time required from the end point of the previous shooting to the initial starting point of the next area after the shooting of the previous area is completed, and the customization of the time from the initial point to the end point of the area.

6. A tissue slice imaging method according to claim 1, characterized in that: Taking the raised blocks of the first tissue slice and the end tissue slice in step three as the dividing points, multiple tissue slices are arranged end to end in sequence, and the specimen slides of the first tissue slice and the end tissue slice are provided with reserved openings so that the raised blocks pass through the reserved openings to ensure that they fit on the specimen slide.

7. A tissue slice imaging method according to claim 6, characterized in that: Based on the relative setting of the raised blocks of the first tissue slice and the end tissue slice, when the first tissue slice is attached to the slide, the image presented by the first tissue slice is a back image, that is, when performing image matching in step three, the image of the first tissue slice is mirror-flipped before image matching.

8. A tissue slice imaging system, comprising an intelligent imaging system, characterized in that: The intelligent imaging system comprises: A two-dimensional imaging unit, used to realize imaging of a single tissue slice; A three-dimensional stitching imaging unit is used to complete three-dimensional visualization imaging by superimposing and stitching multiple two-dimensional images of multiple tissue slices; User interaction interface unit, used to complete the interactive functions of terminal visualization, online human-computer or human-human interaction; A database comparison unit, used for auxiliary diagnosis reference after comparing and analyzing the slice imaging information with the data in the database; Communication unit, completing the communication function between multiple devices; Data acquisition unit, which collects information between multiple devices; A data processing unit, used for analyzing and processing the data information collected by the data acquisition unit; Storage module, used to complete data and system function storage.

Images