A brain tissue slice image sharing system

Abstract

This application provides a brain tissue slice image sharing system, including multiple sharing clients. A first sharing client, responding to a sharing instruction for a target brain tissue slice image, the sharing instruction including at least the target brain tissue slice image and sample data, inputs the target brain tissue slice image into a pre-trained segmentation model to obtain the segmentation result of the target brain tissue slice image output by the segmentation model. The segmentation result is used to indicate key pixel regions and non-key pixel regions in the target brain tissue slice image. The first sharing client encrypts the sample data to generate encrypted sample data. The first sharing client embeds the encrypted sample data into non-key pixel regions according to preset rules to form a processed embedded brain tissue slice image. The first sharing client sends the encrypted embedded brain tissue slice image to a second sharing client indicated by the sharing instruction.

Description

Technical Field

[0001] This application relates to the field of medical data sharing technology, and more specifically, to a system for sharing brain tissue slice images. Background Technology

[0002] In the process of medical digitization, the sharing of medical image data (such as pathology slides and digital slides) across clients and institutions (e.g., remote consultations, research collaborations, teaching image reading) is becoming increasingly frequent. This data typically contains a large amount of patient sample information (such as names, ID numbers, medical record summaries, and diagnostic results), which is highly sensitive. Currently, the conventional technical solutions for protecting patient privacy and data during sharing mainly employ traditional digital watermarking / steganography. Some solutions attempt to use simple LSB (least significant bit) steganography techniques to directly embed information into the image. However, medical slide images have extremely high diagnostic value requirements. Traditional methods blindly embedding data into the entire image introduce unacceptable noise into medical image analysis, affecting doctors' diagnoses or AI-assisted algorithm judgments, and posing medical security risks. Summary of the Invention

[0003] The purpose of this application is to provide a brain tissue slice image sharing system to solve the technical problem in existing medical data sharing technologies that introduce noise into the embedded information of medical image data, affecting subsequent analysis and use.

[0004] In a first aspect, the present invention provides a brain tissue slice image sharing system, comprising multiple sharing clients, wherein a first sharing client responds to a sharing instruction for a target brain tissue slice image, the sharing instruction including at least the target brain tissue slice image and sample data, inputs the target brain tissue slice image into a pre-trained segmentation model to obtain a segmentation result of the target brain tissue slice image output by the segmentation model, the segmentation result being used to indicate key pixel regions and non-key pixel regions in the target brain tissue slice image; the first sharing client encrypts the sample data to generate encrypted sample data; the first sharing client embeds the encrypted sample data into non-key pixel regions according to preset rules to form a processed embedded brain tissue slice image; the first sharing client sends the encrypted embedded brain tissue slice image to a second sharing client indicated by the sharing instruction.

[0005] In an optional implementation, the sharing client generates embedded brain tissue slice images in the following manner: Determine the bytes to be embedded in each pixel within the non-critical pixel region; For each pixel within a non-critical pixel region, the bytes to be embedded are embedded into the target color channel of that pixel to obtain the new pixel value. Based on the new pixel values ​​of pixels in non-critical pixel regions and the pixel values ​​of pixels in critical pixel regions, an embedded brain tissue slice image is generated.

[0006] In an optional implementation, the shared client determines the bytes to be embedded for each pixel within a non-critical pixel region in the following manner: Create a G color channel embedding matrix to allocate bytes in the encrypted sample data to the corresponding pixel positions in non-critical pixel regions; Determine whether the encrypted sample data has been fully allocated; If not, create a B color channel embedding matrix and sequentially allocate the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in the non-critical pixel regions. Determine whether the encrypted sample data has been fully allocated; If not, create an R color channel embedding matrix and sequentially assign the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in the non-critical pixel regions.

[0007] In an optional implementation, the segmentation result also includes the probability value of each pixel in the target brain tissue slice image belonging to a non-critical pixel region. The sharing client writes the bytes to be written from the encrypted sample data into the color channel embedding matrix in the following manner: For each pixel within a non-critical pixel region, the write priority of that pixel is determined based on the probability value that the pixel belongs to the non-critical pixel region. According to the writing priority of each pixel, the non-critical pixel area is divided into multiple pixel sub-regions; The bytes to be written are allocated to each pixel sub-region in descending order of writing priority.

[0008] In an optional implementation, the shared client is also used before the step of determining the write priority of each pixel. The intersection region is determined based on the positional relationship between the relational pixel region and the non-critical pixel region; The intersecting regions are removed from the non-critical pixel regions to form new non-critical pixel regions.

[0009] In an optional implementation, for each pixel within a non-critical pixel region, the sharing client obtains the new pixel value of that pixel in the following manner; Determine the original R channel value, G channel value, and B channel value of the pixel; Perform a bitwise AND operation between the original channel value of the pixel and a preset binary field; Perform a bitwise OR operation between the byte at the same pixel position in the target color channel embedding matrix and the corresponding bitwise AND operation result; Based on the result of the bitwise OR operation, determine the corresponding new channel value; The new pixel value is determined based on the new R channel value, G channel value, and B channel value of the pixel.

[0010] In an optional implementation, the sharing client generates encrypted sample data in the following manner: Convert text-formatted sample data into a binary bitstream; Encrypted sample data is generated by encrypting and compressing a binary bit stream using a preset key.

[0011] In an optional implementation, the first shared client generates a byte embedding key based on the pixel position of the byte to be written, the byte embedding key being used at least to indicate the pixel position and color channel of the byte to be written; The first sharing client sends the encrypted embedded brain tissue slice image, the preset key, and the byte embedding key to the second sharing client as instructed by the sharing command.

[0012] In an optional implementation, the first sharing client performs image quality detection on the embedded brain tissue slice image. If the quality detection passes, the encrypted embedded brain tissue slice image is sent to the second sharing client indicated by the sharing command.

[0013] In an optional implementation, the sharing client performs image quality detection in the following manner: Calculate the peak signal-to-noise ratio and structural similarity index of the embedded brain tissue slice images; If the peak signal-to-noise ratio and structural similarity index of the embedded brain tissue slice image both meet the corresponding threshold conditions, then the quality detection is deemed to have passed.

[0014] This application provides a brain tissue slice image sharing system, comprising multiple sharing clients. A first sharing client, responding to a sharing instruction for a target brain tissue slice image (the instruction includes at least the target brain tissue slice image and sample data), inputs the target brain tissue slice image into a pre-trained segmentation model to obtain the segmentation result of the target brain tissue slice image output by the model. The segmentation result is used to indicate key and non-key pixel regions in the target brain tissue slice image. The first sharing client encrypts the sample data to generate encrypted sample data. The first sharing client embeds the encrypted sample data into non-key pixel regions according to preset rules to form a processed embedded brain tissue slice image. The first sharing client sends the encrypted embedded brain tissue slice image to a second sharing client indicated by the sharing instruction. This sharing system, when clients share brain tissue slice images, intelligently analyzes the slice images to distinguish non-key regions irrelevant to diagnosis and adaptively embeds encrypted sample information into these regions, thereby achieving data fusion at the pixel level while ensuring the integrity of key diagnostic regions of the image. This achieves secure data transmission while guaranteeing the quality of slice image sharing, laying the foundation for subsequent analysis and use. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A schematic diagram of the structure of a brain tissue slice image sharing system provided in an embodiment of this application; Figure 2 A flowchart illustrating the steps of an adaptive steganography method based on morphological features and region selection provided in this application embodiment; Figure 3 This is a flowchart of an information embedding step provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0018] Example 1 Figure 1 This is a schematic diagram of the structure of a brain tissue slice image sharing system provided in an embodiment of this application. Figure 1As shown, this application provides a brain tissue slice image sharing system, which includes multiple sharing clients. These sharing clients can be deployed on electronic devices in the same or different medical institutions, or in the same or different departments.

[0019] Brain tissue samples included in the digital platform of the human brain tissue bank can be used for medical research and teaching. At this time, the slice image and sample information of the target brain tissue slice applied for can be encrypted and sent to the shared client deployed on the application side through the shared client deployed on the management side.

[0020] To protect medical data privacy, existing sharing processes can employ either separate transmission or overall encryption. Overall encryption, using simple LSB steganography, blindly embeds sample information into slice images. However, this often obscures pixel information of key pathological features (such as cell nuclei and abnormal plaques). Even a tiny change in the least significant bit can introduce unacceptable noise into medical image analysis, affecting doctors' diagnoses or AI algorithm judgments, and even posing medical security risks.

[0021] Figure 2 This document provides a flowchart illustrating the steps of an adaptive steganography method based on morphological features and region selection, as illustrated in an embodiment of this application. Specifically, this embodiment provides an adaptive steganography scheme based on morphological features and region selection, which can be implemented through the following steps: S1. The first sharing client responds to the sharing instruction of the target brain tissue slice image. The sharing instruction includes at least the target brain tissue slice image and sample data. The target brain tissue slice image is input into a pre-trained segmentation model to obtain the segmentation result of the target brain tissue slice image output by the segmentation model. The segmentation result is used to indicate the key pixel region and non-key pixel region in the target brain tissue slice image.

[0022] The first and second sharing clients here can be any two different sharing clients in the sharing system. The sharing command can be issued by the requesting sharing client, indicating the second sharing client that needs to share data, the type of request, and the target brain tissue slice image to be requested.

[0023] The sample data here may include the patient ID corresponding to the brain tissue slice image, the slice location (such as coronal / sagittal), the staining type (such as H&E), clinical annotations (such as lesion area markings), etc.

[0024] The shared client here can use a pre-trained segmentation model to perform pixel-to-speech segmentation on brain tissue slice images, identifying key and non-key regions.

[0025] For brain tissue sections, key regions can include morphological structures that are crucial for diagnosis. For example, in neuropathology, structures such as neuronal cell bodies, abnormal protein plaques (e.g., Aβ), and neurofibrillary tangles can be considered key regions. In tumor pathology, cell nuclei, mitotic figures, and abnormal glandular structures can be considered key regions.

[0026] Non-critical areas here can include areas with very low or no diagnostic value. Examples include background areas, interstitial spaces that are not cellular structures, vascular cavities, and cytoplasmic areas that do not stain specifically.

[0027] The segmentation model here takes a brain tissue slice image as input (typically with a resolution of 512×512) and outputs a binary mask of the brain tissue slice image and the probability value of each pixel in the brain tissue slice image belonging to a non-key pixel region. The binary mask is determined based on the probability value of each pixel belonging to either a non-key pixel region or a key pixel region. For example, a probability value of 0.8 can be assigned. If the probability value of a target pixel belonging to a non-key pixel region is greater than 0.8, it belongs to a non-key pixel region, and its pixel value is set to 1; otherwise, it is set to 0, thus generating the binary mask result.

[0028] Furthermore, the segmentation model here can be constructed differently based on the application requirements, such as research discipline. Then, a matching segmentation model can be selected based on the application requirements. The prototype of this segmentation model can be U-Net, nnUNet, or a Transformer-based architecture, etc., without restriction.

[0029] In other embodiments, a pixel-level classification model can be used to obtain the classification object of each pixel, such as background, neurons, or blood vessels. Then, based on the pixel type, it can be determined whether it is a critical or non-critical region.

[0030] S2. The first shared client encrypts the sample data to generate encrypted sample data.

[0031] The sharing client here can compress and encrypt text information in the sample data. For example, the sharing client generates encrypted sample data in the following way: The text-formatted sample data is converted into a binary bitstream. The binary bitstream is then encrypted and compressed using a preset key to generate encrypted sample data.

[0032] In a specific implementation, the AES-256-CBC algorithm can be used for encryption, outputting a fixed byte sequence. The resulting hexadecimal characters are then converted into binary encrypted sample data. The encryption key can be distributed uniformly by the management side.

[0033] Other encryption algorithms can also be used here, without limitation.

[0034] S3. The first shared client embeds the encrypted sample data into non-critical pixel regions according to preset rules to form a processed embedded brain tissue slice image.

[0035] Figure 3 A flowchart illustrating an information embedding step provided in this application is shown. In step S3, the sharing client generates an embedded brain tissue slice image in the following manner: S30. Determine the bytes to be embedded in each pixel within the non-critical pixel region.

[0036] In step S30, the shared client can determine the bytes to be embedded for each pixel within the non-critical pixel region in the following way: Create a G color channel embedding matrix to allocate bytes in the encrypted sample data to the corresponding pixel positions in non-critical pixel regions; determine if the encrypted sample data is fully allocated; if not, create a B color channel embedding matrix to allocate the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in non-critical pixel regions in sequence; determine if the encrypted sample data is fully allocated; if not, create an R color channel embedding matrix to allocate the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in non-critical pixel regions in sequence.

[0037] For brain tissue slice images generated using HE staining, in order to reduce the impact of embedded information on image quality, information can be embedded first into the G and B color channels, and finally into the R color channel.

[0038] Here, the color channel embedding matrix indicates the bytes to be embedded for the corresponding color channel at each pixel location. The color channel embedding matrix is ​​set to correspond to the resolution of the brain tissue slice image. For example, for a 512×512 brain tissue slice image, the corresponding color channel embedding matrix is ​​also 512×512 in dimension.

[0039] The shared client writes the bytes to be written from the encrypted sample data into the color channel embedding matrix in the following way: For each pixel within a non-critical pixel region, its write priority is determined based on the probability value of that pixel belonging to the non-critical pixel region. According to the write priority of each pixel, the non-critical pixel region is divided into multiple pixel sub-regions. The bytes to be written are then sequentially allocated to each pixel sub-region in descending order of write priority. These pixel sub-regions are not necessarily connected.

[0040] In specific implementations, write priorities can be divided into three levels from high to low: Level 1, Level 2, and Level 3, based on probability values.

[0041] For example, encrypted sample data , This represents the total number of bytes.

[0042] For the pixel sub-region corresponding to the first write priority, the corresponding to The pixels are assigned to corresponding pixel positions in order from top to bottom and from left to right. This represents the total number of pixels in the pixel sub-region corresponding to the first write priority.

[0043] If the encrypted sample data is not completely allocated, for the pixel sub-region corresponding to the second write priority, it can be... to Distribute in order. This represents the total number of pixels in the sub-region corresponding to the first write priority. If the encrypted sample data has not been completely allocated, then for the sub-region corresponding to the third write priority, the number of pixels can be increased. to Allocation. At this point, the pixel sub-region corresponding to the third write priority may not be fully allocated.

[0044] It should be noted that unassigned pixel positions are left blank. The order in which bytes are allocated within each sub-region can also be different; this is not limited here. However, when parsing steganographic data, a reading order with the same allocation order is required to combine and generate encrypted sample data.

[0045] In this way, prioritizing the writing of non-critical pixels with low diagnostic decision weight and color channels can reduce the impact on image quality.

[0046] S31. For each pixel in the non-critical pixel region, embed the bytes to be embedded into the target color channel of that pixel to obtain the new pixel value.

[0047] In step S31, for each pixel within a non-critical pixel region, the sharing client can obtain the new pixel value of that pixel in the following manner; Determine the original R, G, and B channel values ​​of the pixel. Perform a bitwise AND operation on the original channel values ​​of the pixel with a preset binary field. Perform a bitwise OR operation on the byte at the same pixel position in the target color channel embedding matrix with the corresponding bitwise AND result. Determine the corresponding new channel value based on the bitwise OR result. Determine the new pixel value based on the new R, G, and B channel values ​​of the pixel.

[0048] Here, in pixels For example, the RGB value is [r=80, g=170, b=120]. The value at the corresponding pixel position of the G color channel embedding matrix is ​​bit1=1, the value at the corresponding pixel position of the B color channel embedding matrix is ​​bit2=0, and the value at the corresponding pixel position of the R color channel embedding matrix is ​​empty.

[0049] The new pixel value g' = (g & 0xFE) | bit1 = 10101011 (binary) = 171 (decimal), where g & 0xFE is used to clear the least significant bit of g, & represents bitwise operation, and | bit1 is used to embed bit1 into the least significant bit of g. Similarly, b' = (b & 0xFE) | bit2 = 01111000 (binary) = 120 (decimal).

[0050] Therefore, the new RGB values ​​are [r=80, g=171, b=120].

[0051] In one embodiment of this application, in order to further blur the brain tissue slice image to meet the requirements of secure data sharing, the 0 and 1 bytes to be written can be transformed, for example, 0 can be replaced with 5 and 1 can be replaced with 9, and the transformed value can be embedded in the least significant bit.

[0052] In one embodiment of this application, in addition to embedding the least significant bit, the most significant bit can also be embedded. This not only achieves the embedding of steganographic information, but also further blurs the brain tissue slice image, so that even if the embedded brain tissue slice image is cracked, the real slice data will not be exposed.

[0053] S32. Generate an embedded brain tissue slice image based on the new pixel values ​​of pixels in non-key pixel regions and the pixel values ​​of pixels in key pixel regions.

[0054] The resulting fused image of the embedded brain tissue slice is visually almost indistinguishable from the original image. It can be opened by any standard image viewer and used for diagnosis or other purposes, while also carrying complete encrypted sample data.

[0055] S4. The first sharing client sends the encrypted embedded brain tissue slice image to the second sharing client as instructed by the sharing command.

[0056] There are no restrictions on the encryption method for sliced ​​images here.

[0057] This application provides a brain tissue slice image sharing system that, through intelligent region selection, avoids embedding data in key pathological features, thus ensuring the diagnostic quality of the images. Simultaneously, the sample data is encrypted before embedding, and the embedding process itself is covert, further enhancing security.

[0058] Example 2 In one embodiment of this application, prior to step S30, the shared client may further determine the intersection region based on the positional relationship between the key pixel region and the non-key pixel region, before determining the write priority of each pixel. The intersection region is then removed from the non-key pixel region to become a new non-key pixel region.

[0059] Specifically, the intersection area can be determined based on the proportion of key pixels and non-key pixels within a preset-sized pixel window. If the proportion of key pixels is greater than 80%, it can be determined as an intersection area.

[0060] For some high-precision analysis needs, to avoid the cross-influence of non-critical region data steganography on pixels in adjacent critical regions, sensitive regions that highly overlap with critical regions can be removed and excluded from the embedded pixel objects. Furthermore, since non-critical regions typically constitute a large proportion of sliced ​​images, this step will not affect the integrity of the embedded data.

[0061] Example 3 In one embodiment of this application, in some high-precision analysis scenarios, The first sharing client performs image quality checks on the embedded brain tissue slice image. If the quality check passes, it sends the encrypted embedded brain tissue slice image to the second sharing client specified in the sharing command. The sharing client can perform image quality checks in the following ways: Calculate the peak signal-to-noise ratio (PSNR) and structural similarity index of the embedded brain tissue slice image. If both the PSNR and structural similarity index of the embedded brain tissue slice image meet the corresponding threshold conditions, the quality detection is deemed successful.

[0062] Before sharing brain tissue slice images, the embedded brain tissue slice images with embedded sample information can be subjected to image quality assessment. If the quality test is passed, the slice images can be shared. If the quality test is not passed, the system can consider using other sharing schemes for encryption and transmission.

[0063] Example 4 In one embodiment of this application, the first sharing client may also embed a short piece of metadata in a standard LSB manner at a fixed position (such as a few specific pixels on the edge of the image) of the sliced ​​image to identify whether the image contains hidden information, the algorithm version used, the data length, etc., so as to facilitate the receiver to identify and extract it.

[0064] Example 5 In a feasible embodiment, the first sharing client can generate a byte embedding key based on the pixel positions allocated to the bytes to be written. The byte embedding key is used to indicate at least the pixel position and color channel of the bytes to be written, and may also include the writing priority corresponding to each pixel position. The first sharing client sends the encrypted embedded brain tissue slice image, the preset key, and the byte embedding key to the second sharing client indicated by the sharing instruction.

[0065] The shared client can determine the byte embedding key based on the color channel embedding matrix. The second shared client can then extract the corresponding encrypted sample data at the location embedded in the brain tissue slice image based on this byte embedding key, where 0 to 3 bytes can be extracted from each pixel location. Furthermore, the second shared client can also reconstruct the embedded brain tissue slice image based on the bytes extracted from each pixel location.

[0066] The second sharing client can obtain the complete sample data by decoding the encrypted sample data based on the preset key.

[0067] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0068] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0069] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0070] It should be noted that if the function is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0071] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0072] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A system for sharing brain tissue slice images, characterized in that, Includes multiple shared clients, among which, The first sharing client responds to the sharing instruction of the target brain tissue slice image. The sharing instruction includes at least the target brain tissue slice image and sample data. The target brain tissue slice image is input into a pre-trained segmentation model to obtain the segmentation result of the target brain tissue slice image output by the segmentation model. The segmentation result is used to indicate the key pixel region and non-key pixel region in the target brain tissue slice image. The first sharing client encrypts the sample data to generate encrypted sample data; The first shared client embeds encrypted sample data into non-critical pixel regions according to preset rules to form processed embedded brain tissue slice images; The first sharing client sends the encrypted embedded brain tissue slice image to the second sharing client indicated by the sharing command.

2. The system according to claim 1, characterized in that, The shared client generates embedded brain tissue slice images in the following ways: Determine the bytes to be embedded in each pixel within the non-critical pixel region; For each pixel within a non-critical pixel region, the bytes to be embedded are embedded into the target color channel of that pixel to obtain the new pixel value. Based on the new pixel values ​​of pixels in non-critical pixel regions and the pixel values ​​of pixels in critical pixel regions, an embedded brain tissue slice image is generated.

3. The system according to claim 2, characterized in that, The shared client determines the bytes to be embedded for each pixel within a non-critical pixel region using the following method: Create a G color channel embedding matrix to allocate bytes in the encrypted sample data to the corresponding pixel positions in non-critical pixel regions; Determine whether the encrypted sample data has been fully allocated; If not, create a B color channel embedding matrix and sequentially allocate the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in the non-critical pixel regions. Determine whether the encrypted sample data has been fully allocated; If not, create an R color channel embedding matrix and sequentially assign the unallocated bytes in the current encrypted sample data to the corresponding pixel positions in the non-critical pixel regions.

4. The system according to claim 3, characterized in that, The segmentation result also includes the probability value of each pixel in the target brain tissue slice image belonging to a non-critical pixel region. The shared client writes the bytes to be written from the encrypted sample data into the color channel embedding matrix in the following way: For each pixel within a non-critical pixel region, the write priority of that pixel is determined based on the probability value that the pixel belongs to the non-critical pixel region. According to the writing priority of each pixel, the non-critical pixel area is divided into multiple pixel sub-regions; The bytes to be written are allocated to each pixel sub-region in descending order of writing priority.

5. The system according to claim 4, characterized in that, The shared client is also used before the step of determining the write priority of each pixel. The intersection region is determined based on the positional relationship between the relational pixel region and the non-critical pixel region; The intersecting regions are removed from the non-critical pixel regions to form new non-critical pixel regions.

6. The system according to claim 3, characterized in that, For each pixel within a non-critical pixel region, the sharing client obtains the new pixel value of that pixel in the following manner; Determine the original R channel value, G channel value, and B channel value of the pixel; Perform a bitwise AND operation between the original channel value of the pixel and a preset binary field; Perform a bitwise OR operation between the byte at the same pixel position in the target color channel embedding matrix and the corresponding bitwise AND operation result; Based on the result of the bitwise OR operation, the corresponding new channel value is determined; The new pixel value is determined based on the new R channel value, G channel value, and B channel value of the pixel.

7. The system according to claim 4, characterized in that, The shared client generates encrypted sample data in the following ways: Convert text-formatted sample data into a binary bitstream; Encrypted sample data is generated by encrypting and compressing a binary bit stream using a preset key.

8. The system according to claim 7, characterized in that, The first shared client generates a byte embedding key based on the pixel position of the byte to be written, the byte embedding key being used at least to indicate the pixel position and color channel of the byte to be written; The first sharing client sends the encrypted embedded brain tissue slice image, the preset key, and the byte embedding key to the second sharing client indicated by the sharing instruction.

9. The system according to claim 1, characterized in that, The first sharing client performs image quality detection on the embedded brain tissue slice image. If the quality detection passes, the encrypted embedded brain tissue slice image is sent to the second sharing client indicated by the sharing command.

10. The system according to claim 9, characterized in that, The shared client performs image quality detection in the following ways: Calculate the peak signal-to-noise ratio and structural similarity index of the embedded brain tissue slice images; If the peak signal-to-noise ratio and structural similarity index of the embedded brain tissue slice image both meet the corresponding threshold conditions, then the quality detection is deemed to have passed.

Images