An interactive region of interest based medical image remote transmission method and device

Abstract

The application provides a medical image remote transmission method and device based on an interactive region of interest. The method comprises the following steps: a server determines an initial ROI region according to a segmentation result and feature points; the server generates an original sampling rate mask according to the initial ROI region, and generates a low sampling rate mask for a whole image; data transmission based on the mask; use of the data by a client; expansion of the ROI region at the client; transmission of newly generated expansion ROI region description information to the server; extraction of an original sampling rate image by the server and return of the original sampling rate image to the client; and rendering of the received original sampling rate image by the client, and modification of the original sampling rate mask and the low sampling rate mask. The method provided by the application realizes more efficient and faster data transmission in optimizing medical image remote transmission, improves the sampling rate of a key region according to the will of a doctor, improves the diagnosis ability of the doctor for a patient image, and reduces the occupation of network resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical image processing, and particularly to a medical image remote transmission method and device based on an interactive region of interest. BACKGROUND

[0002] In the transmission of CT, MRI and other medical data (mainly volume data) on the Internet, the efficiency of medical data transmission is relatively low. Conventional medical image data has a huge volume, which brings great pressure to Internet remote transmission, easily causes network congestion, occupies a large amount of network resources, affects the transmission and communication of other important data, directly affects the real-time interaction and actual experience of doctors with image data, and the time for doctors to wait for data download to complete is very long. At the same time, due to the overall transmission, the data volume of medical images generated for different surgical procedures and cases cannot be intelligently adjusted according to actual needs, wasting network resources and storage space.

[0003] Patent CN104134214A provides a digital slice image storage and display method based on image pyramid layering, which splits a piece of image data in advance according to multiple levels of resolution / sampling rate, and the server returns the resolution / sampling rate image of the corresponding area on demand when browsing in the front end. However, this method has the following problems:

[0004] (1) A plurality of levels of resolution / sampling rate image tiles need to be prepared in advance, wasting a large amount of storage space;

[0005] (2) Spaces such as patient pallets, device edges, and surrounding air that have no diagnostic significance are also sliced at multiple levels, wasting a large amount of storage space;

[0006] (3) Transmission can only be in the form of a large number of tiles, and the data contained in these tiles is relatively small, so the compression efficiency is very low when applying image compression and general compression algorithms. Therefore, the pyramid storage structure scheme in most scenarios is transmitted in bitmap mode, which is not suitable for any compression technology. SUMMARY

[0007] The present application aims to solve the shortcomings of the prior art described in the background art, and provides a medical image remote transmission method and device based on an interactive region of interest. The method optimizes medical image remote transmission, achieves more efficient and faster data transmission, and increases the sampling rate of key areas according to the wishes of doctors, improves the diagnostic ability of doctors for patient images, reduces the occupation of network resources, and provides a more feasible solution for remote medical treatment.

[0008] The application is implemented by the following technical solutions: in a first aspect, the application provides a medical image remote transmission method based on an interactive region of interest, comprising the following steps:

[0009] S101, the server determines an initial ROI region according to the segmentation result and the feature point position;

[0010] S102, the server generates an original sampling rate mask according to the initial ROI region, and generates a low sampling rate mask for the whole image; wherein the mask refers to a binary volume data completely coinciding with the original volume data space position; 1 represents the current need for rendering, and 0 represents the current no need for rendering;

[0011] S103, data transmission based on the mask, comprising the following steps:

[0012] The server generates an original sampling rate data stream according to the original sampling rate mask, and extracts the data within the ROI as a voxel data stream;

[0013] The original sampling rate data stream, the voxel data stream, the data after overall downsampling, the ROI low sampling rate mask and the original sampling rate mask are sent to the client;

[0014] Among them, the data transmission is all compressed transmission, and since the mask data is a 01 byte stream, a very high compression ratio can be achieved;

[0015] S104, the use of data by the client, comprising the following two modes:

[0016] The client restores the complete original image data: the interpolation is performed on the down-sampled image data, the voxel interval of the original data is restored, and new image data V is generated; according to the original sampling rate mask, the corresponding voxel in the image data V is replaced by the value in the voxel data stream;

[0017] The client simultaneously displays the original sampling rate image data of the ROI region in space, the data after downsampling of the application of the downsampling mask, and performs mixed display based on the multi-volume data; wherein the original sampling rate image data is formed by filling the voxel stream according to the original resolution mask; the mixed display supports slice display and volume rendering;

[0018] S105, the extension ROI region is adjusted by adjusting the feature point position or directly outlined on the slice at the client side;

[0019] S106, the newly generated extension ROI region description information is sent to the server, and the server extracts the original sampling rate image according to the received extension ROI region and returns it to the client;

[0020] S107, the client renders the received original sampling rate image, and modifies the original sampling rate mask and the low sampling rate mask;

[0021] Original sampling rate mask merge extended ROI region, i.e. voxel replacement to the binarized mask;

[0022] Low sampling rate mask subtract extended ROI region, i.e. voxel replacement to the binarized mask;

[0023] S108, the rendering of the images of two sampling rates on the client side through the mask is completed.

[0024] Further, the server in S101 determines the initial ROI region according to the segmentation result and the feature point position, and the following cases are divided according to different procedures:

[0025] Hip replacement surgery: the region of interest of the hip bone is defined as starting from the hip bone model vertex and extending outward along the vertex normal vector by a distance A; the proximal femur region of interest is defined as passing through the femoral small trochanter feature point and perpendicular to the plane above the part of the vertex of the femoral mechanical axis, and extending outward along the vertex normal vector by a distance A; the femoral knee joint region of interest is defined as starting from the femoral knee joint model vertex and extending outward along the vertex normal vector by a distance A; for a specific anatomical feature point, the region of interest is defined as the feature point position as the center of a sphere, and the region of interest is extended outward by a distance B; the total region of interest is the union of all the above regions of interest;

[0026] Total knee arthroplasty, unicompartmental knee arthroplasty: the region of interest of the proximal femur is defined as the femoral head center feature point outwardly extended by a distance A; the region of interest of the distal femur is defined as starting from the femoral side knee joint model vertex and extending outward along the vertex normal vector by a distance A; the proximal tibial region of interest is defined as starting from the tibial side knee joint model vertex and extending outward along the vertex normal vector by a distance A; the distal tibial region of interest is defined as starting from the ankle joint model vertex and extending outward along the vertex normal vector by a distance A; for a specific anatomical feature point, the region of interest is defined as the feature point position as the center of a sphere, and the region of interest is extended outward by a distance B; the total region of interest is the union of all the above regions of interest;

[0027] Other procedures: the bone region of interest is defined as starting from the bone model vertex and extending outward along the vertex normal vector by a distance A; for a specific anatomical feature point, the region of interest is defined as the feature point position as the center of a sphere, and the region of interest is extended outward by a distance B; the total region of interest is the union of all the above regions of interest.

[0028] Further, the steps in S102 also include:

[0029] Aligning the high sampling rate mask and the low sampling rate mask: the edge length of the high sampling rate mask is an integer multiple of the low sampling rate voxel interval.

[0030] Further, the extended ROI region in S105 on the client side is adjusted by adjusting the feature point position or directly outlining on the slice, including:

[0031] By moving the point: by moving the specific feature points on the image, based on the moving interaction, these feature points will expand the distance B as the expanded ROI region;

[0032] Manual delineation of the region: by manually framing a new region of interest on the image, the system will automatically expand the ROI according to the manually delineated position and range; the expanded ROI region is the delineated region and the region expanded by the thickness C in the up and down directions based on the slice.

[0033] In a second aspect, the present application also provides a medical image remote transmission device based on an interactive region of interest, which comprises:

[0034] A module for determining an initial ROI region, wherein the server determines the initial ROI region according to the segmentation result and the feature point position;

[0035] A module for generating a mask, wherein the server generates a raw sampling rate mask according to the initial ROI region and a low sampling rate mask for the whole image; wherein the mask refers to a binary volume data completely coinciding with the spatial position of the raw volume data; 1 represents the current need for rendering, and 0 represents the current no need for rendering;

[0036] A data transmission module based on the mask, comprising the following steps:

[0037] The server generates a raw sampling rate data stream according to the raw sampling rate mask, and the data within the ROI is extracted as a voxel data stream;

[0038] The raw sampling rate data stream, the voxel data stream, the data after the whole downsampling, the ROI low sampling rate mask, and the raw sampling rate mask are sent to the client;

[0039] Wherein, during data transmission, compression transmission is adopted, and since the mask data is a 01 byte stream, a very high compression ratio can be achieved;

[0040] A data usage module, including the following two modes:

[0041] The client restores the complete raw image data: the interpolation is performed on the down-sampled image data to restore the voxel interval of the original data, and new image data V is generated; according to the raw sampling rate mask, the corresponding voxel in the image data V is replaced by the value in the voxel data stream;

[0042] The client simultaneously displays the raw sampling rate image data of the ROI region in space, the down-sampled data after applying the down-sampling mask, and performs mixed display based on the multi-volume data; wherein the raw sampling rate image data is formed by filling the voxel stream according to the original resolution mask; the mixed display supports slice display and volume rendering;

[0043] An extended ROI region module, which is used to adjust the feature points or directly draw the extended ROI region on the slice at the client side;

[0044] A sending and receiving module, which is used to send the extended ROI region description information to the server, and the server extracts the original sampling rate image according to the received extended ROI region and returns it to the client;

[0045] A rendering module, which is used to render the received original sampling rate image, and modify the original sampling rate mask and the low sampling rate mask at the client side;

[0046] The original sampling rate mask is combined with the extended ROI region, that is, the binary mask is replaced by voxel;

[0047] The low sampling rate mask is subtracted from the extended ROI region, that is, the binary mask is replaced by voxel;

[0048] An image rendering module, which is used to render the images of two sampling rates through the mask at the client side.

[0049] The present application provides a medical image remote transmission method and device based on an interactive region of interest.

[0050] 1. By using the interactive ROI setting and control, the fine sampling rate control of the medical image is realized, the key region is kept at a high sampling rate, and the non-key region is down-sampled, so that the data amount of transmission is significantly reduced;

[0051] 2. The voxels can be compressed and transmitted in different sampling regions, and the general compression algorithm can be applied to various images to obtain a good compression rate and further improve the transmission efficiency;

[0052] 3. The interactive design is introduced, the doctor can modify the region of interest according to the needs, realize the customized operation of the image, improve the interaction between the doctor and the remote patient data, and enhance the flexibility in the remote medical data transmission scene;

[0053] 4. Based on the interactive ROI, the server can combine and extract data in real time as needed, avoiding unnecessary waste of network resources;

[0054] 5. The mask technology and mask alignment scheme are used to realize the natural transition of images with different sampling rates, and ensure that the images presented at the client side are seamlessly spliced in and out of the key region. BRIEF DESCRIPTION OF DRAWINGS

[0055] The features, advantages and technical effects of the exemplary embodiments of the present application will be described below with reference to the accompanying drawings.

[0056] Figure 1A medical image remote transmission method based on an interactive region of interest is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0057] For the purpose of making the purpose, technical scheme and advantages of the present disclosure clearer, the features and exemplary embodiments of various aspects of the present disclosure will be described in detail below. It should be understood that the specific embodiments described herein are only intended to explain the present disclosure, rather than limit the present disclosure. The present disclosure can be implemented without some of these specific details by those skilled in the art. The following description of the embodiments is only to provide a better understanding of the present disclosure by showing examples of the present disclosure.

[0058] It should be noted that, in this paper, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment. Without more limitations, the elements defined by the statement "include" do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0059] In order to better understand the present application, the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0060] As shown in Figure 1 The present application provides a medical image remote transmission method based on an interactive region of interest, comprising the following steps:

[0061] S101, the server determines the initial ROI region according to the segmentation result and the feature point position;

[0062] S102, the server generates the original sampling rate mask according to the initial ROI region, and generates the low sampling rate mask for the whole image; wherein the mask refers to the binary volume data completely coinciding with the original volume data space position; 1 represents the current rendering, and 0 represents the current non-rendering;

[0063] S103, data transmission based on the mask, comprising the following steps:

[0064] The server generates the original sampling rate data stream according to the original sampling rate mask, and extracts the data within the ROI as a voxel data stream;

[0065] The original sampling rate data stream, the voxel data stream, the overall down-sampled data, the ROI low sampling rate mask, and the original sampling rate mask are sent to the client;

[0066] Among them, the data transmission is compressed transmission, and since the mask data is a 01 byte stream, it can achieve a very high compression ratio.

[0067] S104, the client uses the data, including the following two ways:

[0068] The client restores the complete original image data: the down-sampled image data is interpolated to restore the voxel interval of the original data, and new image data V is generated; according to the original sampling rate mask, the corresponding voxel in the image data V is replaced by taking the value in the voxel data stream;

[0069] The client simultaneously displays the original sampling rate image data of the ROI region in space, the down-sampled data of the application down-sampling mask, and the mixed display based on multi-body data; wherein the original sampling rate image data is formed by filling the voxel stream according to the original resolution mask; the mixed display supports slice display and volume rendering;

[0070] S105, the client adjusts the feature point position or directly outlines the expanded ROI region on the slice;

[0071] S106, the newly generated expanded ROI region description information is sent to the server, and the server extracts the original sampling rate image according to the received expanded ROI region and returns it to the client;

[0072] S107, the client renders the received original sampling rate image, and modifies the original sampling rate mask and the low sampling rate mask;

[0073] The original sampling rate mask merges the expanded ROI region, that is, the binary mask is replaced by voxel;

[0074] The low sampling rate mask subtracts the expanded ROI region, that is, the binary mask is replaced by voxel;

[0075] S108, the rendering of the images of the two sampling rates in the client through the mask is completed.

[0076] As an optional implementation, the server in S101 determines the initial ROI region according to the segmentation result and the feature point position, and according to different surgical procedures, it can be divided into the following cases:

[0077] Hip replacement surgery: the region of interest of the hip bone is defined as a sphere with a center at the top point of the hip bone model and a radius A; the region of interest of the proximal femur is defined as a sphere with a center at the top point of the proximal femur model and a radius A; the region of interest of the knee joint of the femur is defined as a sphere with a center at the top point of the knee joint model of the femur and a radius A; for a specific anatomical feature point, the region of interest is defined as a sphere with a center at the feature point and a radius B; the total region of interest is the union of all the above regions of interest;

[0078] Total knee arthroplasty, unicompartmental knee arthroplasty: the region of interest of the proximal femur is defined as a sphere with a center at the top point of the proximal femur model and a radius A; the region of interest of the distal femur is defined as a sphere with a center at the top point of the knee joint model of the femur and a radius A; the region of interest of the proximal tibia is defined as a sphere with a center at the top point of the knee joint model of the tibia and a radius A; the region of interest of the distal tibia is defined as a sphere with a center at the top point of the ankle joint model and a radius A; for a specific anatomical feature point, the region of interest is defined as a sphere with a center at the feature point and a radius B; the total region of interest is the union of all the above regions of interest;

[0079] Other procedures: the region of interest of the bone is defined as a sphere with a center at the top point of the bone model and a radius A; for a specific anatomical feature point, the region of interest is defined as a sphere with a center at the feature point and a radius B; the total region of interest is the union of all the above regions of interest.

[0080] Typically, A is 15 cm and B is 10 cm.

[0081] As an optional implementation, the step in S102 further includes:

[0082] Aligning the high sampling rate mask and the low sampling rate mask: the edge length of the high sampling rate mask is an integer multiple of the low sampling rate voxel interval.

[0083] As an optional implementation, the step of adjusting the feature point position or directly outlining the extended ROI region on the slice in S105 includes:

[0084] By moving the point: by moving a specific feature point, such as the center of the medullary cavity, the center of the knee joint, etc., on the image, based on the moving interaction, the feature point will expand outwardly by a sphere with a radius B as the extended ROI region;

[0085] Manual delineation region: a new region of interest is manually delineated on the image by manual framing, and the system will automatically expand the ROI according to the position and range of the manual delineation; the expanded ROI region is the delineated region and the region expanded by a thickness C above and below the vertical direction of the slice.

[0086] Generally, B is 10 centimeters, and C is 10 centimeters.

[0087] In a second aspect, the present application also provides a medical image remote transmission device based on an interactive region of interest, which comprises:

[0088] An initial ROI region determination module, in which the server determines an initial ROI region according to the segmentation result and the feature point position;

[0089] A mask generation module, in which the server generates an original sampling rate mask according to the initial ROI region and generates a low sampling rate mask for the whole image; wherein the mask refers to a binary volume data completely coinciding with the original volume data space position; 1 represents the current need for rendering, and 0 represents the current no need for rendering;

[0090] A data transmission module based on the mask, comprising the following steps:

[0091] The server generates an original sampling rate data stream according to the original sampling rate mask, and the data within the ROI is extracted as a voxel data stream;

[0092] The original sampling rate data stream, the voxel data stream, the data after overall downsampling, the ROI low sampling rate mask, and the original sampling rate mask are sent to the client;

[0093] Wherein, during data transmission, compression transmission is adopted, and since the mask data is a 01 byte stream, a very high compression ratio can be achieved;

[0094] A data usage module, in which the client uses the data in the following two ways:

[0095] The client restores the complete original image data: the down-sampled image data is interpolated to restore the voxel interval of the original data, and new image data V is generated; according to the original sampling rate mask, the corresponding voxels in the image data V are replaced by the values in the voxel data stream;

[0096] The client simultaneously displays the original sampling rate image data of the ROI region in space, the data after applying the down-sampling mask, and the mixed display based on multi-volume data; wherein the original sampling rate image data is formed by filling the voxel stream according to the original resolution mask; the mixed display supports slice display and volume rendering;

[0097] An expanded ROI region module, in which the client adjusts the feature point position or directly delineates the expanded ROI region on the slice;

[0098] The sending-receiving module sends the newly generated extended ROI region description information to the server, and the server extracts the original sampling rate image according to the received extended ROI region and returns it to the client;

[0099] The rendering module renders the received original sampling rate image, and modifies the original sampling rate mask and the low sampling rate mask;

[0100] The original sampling rate mask is merged with the extended ROI region, that is, the binary mask is voxel-replaced;

[0101] The low sampling rate mask is subtracted from the extended ROI region, that is, the binary mask is voxel-replaced;

[0102] The image rendering module completes the rendering of the images of the two sampling rates on the client side through the mask.

[0103] Each module / unit in the device has the function of realizing each step in the medical image remote transmission method based on the interactive region of interest, and can achieve its corresponding technical effects. For the sake of brevity, no further description is given here.

[0104] The present application provides a medical image remote transmission method and device based on an interactive region of interest. Compared with the prior art, the present application has the following technical advantages:

[0105] 1. By adopting interactive ROI setting and control, fine sampling rate control of medical images is realized, the key region is kept at a high sampling rate, and the non-key region is down-sampled, thereby significantly reducing the data amount of transmission;

[0106] 2. Different sampling regions can transmit voxel streams in a compressed manner, and general compression algorithms can be applied to various images to obtain a good compression rate and further improve the transmission efficiency;

[0107] 3. The interactive design is introduced, the doctor can modify the region of interest according to the need, realizes the customized operation of the image, improves the interactivity of the doctor and the remote patient data, and enhances the flexibility in the remote medical data transmission scene;

[0108] 4. Based on the interactive ROI, the server can combine and extract data in real time as needed, avoiding unnecessary waste of network resources;

[0109] 5. The mask technology and mask alignment scheme are used to realize the natural transition of images with different sampling rates, and ensure that the images presented on the client side are seamlessly spliced in and out of the key region.

[0110] It is to be expressly understood that the disclosure is not limited to the particular configurations, process steps, and materials described and illustrated herein, as such can vary. Detailed

[0111] The functions illustrated in the structural block diagrams described above can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, and the like. When implemented in software, the elements of the disclosure are program or code segments used to perform the required tasks. The relevant skilled person can write the computer program code for performing the operations of the present disclosure in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" language or similar programming languages. In addition, the program or code segment can be stored in a machine-readable medium or transmitted over a transmission medium or communication link by a data signal carried in a carrier wave. The machine-readable medium can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, and the like.

[0112] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks. Alternatively, computer program implemented processes can be produced on less programmable apparatus by utilizing express instructions to cause a series of operational steps to be performed on the less programmable apparatus to produce the computer implemented process such that the instructions which execute on the less programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0113] The above description is only a specific implementation of the present disclosure. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, modules and units can refer to the corresponding processes in the foregoing method embodiments, which will not be described herein. It should be understood that the protection scope of the present disclosure is not limited in this way, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present disclosure, and these modifications or replacements should be covered by the protection scope of the present disclosure.

Claims

1. A method for remote transmission of medical images based on interactive regions of interest, characterized in that, Includes the following steps: S101, The server determines the initial ROI region based on the segmentation results and feature points; S102, the server generates the original sampling rate mask based on the initial ROI region, and generates a low sampling rate mask for the ROI across the entire image; where the mask refers to the binarized volume data that completely overlaps with the spatial location of the original volume data; 1 represents that rendering is required at the moment, and 0 represents that rendering is not required at the moment; S103, Mask-based data transmission includes the following steps: The server generates a raw sampling rate data stream based on the raw sampling rate mask, and extracts the data within the ROI as a voxel data stream. Send the original sampling rate data stream, voxel data stream, downsampled image data, ROI low sampling rate mask, and original sampling rate mask to the client; All data transmissions are compressed, and since the mask data is a 01-byte stream, an extremely high compression ratio can be achieved. S104, The client's use of data includes the following two methods: The client restores the complete original image data: interpolates the downsampled image data to restore the voxel interval of the original data and generates new image data V; based on the original sampling rate mask, it takes values ​​from the voxel data stream to replace the corresponding voxels in the image data V; The client simultaneously displays the original sampling rate image data and downsampled image data of the ROI region in space, and performs a mixed display based on multi-volume data; the original sampling rate image data is formed by filling the voxel stream according to the original sampling rate mask; the mixed display supports slice display and volume rendering; S105, on the client side, the ROI region can be extended by adjusting the feature points or by directly drawing on the slice; S106, The newly generated extended ROI region description information is sent to the server. The server extracts the original sampling rate image based on the received extended ROI region and returns it to the client. S107, the client renders the received original sampling rate image and modifies the original sampling rate mask and the ROI low sampling rate mask; The original sampling rate mask is merged and the ROI region is expanded, which means voxel replacement is performed on the binarized mask; The low sampling rate ROI mask is subtracted from the extended ROI region, which is equivalent to voxel replacement of the binarized mask; S108 completes the rendering of images at two sampling rates on the client side via a mask.

2. The method for remote transmission of medical images based on interactive regions of interest according to claim 1, characterized in that, The server described in S101 determines the initial ROI region based on the segmentation results and feature points, and the determination can be categorized into the following cases depending on the technique used: Hip replacement surgery: The region of interest (ROI) for the hip bone is defined as starting from the vertex of the hip bone model and extending outwards by a distance A along the vertex's normal vector; the proximal femoral ROI is defined as the portion of the vertices above the lesser trochanter feature point and perpendicular to the femoral mechanical axis plane, extending outwards by a distance A along the vertex's normal vector; the femoral-knee joint region ROI is defined as starting from the vertex of the femoral-knee joint model and extending outwards by a distance A along the vertex's normal vector; for a specific anatomical feature point, the ROI is defined as extending outwards by a distance B with the feature point as the center of the sphere; the total ROI is the union of all the above ROIs. Total knee arthroplasty and unicompartmental knee arthroplasty: The region of interest (ROI) for the proximal femur is defined as a spherical extension A from the central feature point of the femoral head; the region of interest for the distal femur is defined as a distance A extending outward from the apex of the femoral-side knee joint model along the normal vector of the apex; the region of interest for the proximal tibia is defined as a distance A extending outward from the apex of the tibial-side knee joint model along the normal vector of the apex; the region of interest for the distal tibia is defined as a distance A extending outward from the apex of the ankle joint model along the normal vector of the apex; for a specific anatomical feature point, the ROI is defined as a distance B extending outward from the feature point as the center of a sphere; the total ROI is the union of all the above ROIs. Other techniques: The region of interest for a skeleton is defined as extending outwards by a distance A from the vertex of the skeleton model along the vertex normal vector; for a specific anatomical feature point, the region of interest is defined as extending outwards by a distance B with the feature point as the center of the sphere; the total region of interest is the union of all the above regions of interest.

3. The method for remote transmission of medical images based on interactive regions of interest according to claim 1, characterized in that, The steps in S102 also include: Align the original sampling rate mask and the ROI low sampling rate mask: the edge length of the original sampling rate mask is an integer multiple of the low sampling rate voxel interval.

4. The method for remote transmission of medical images based on interactive regions of interest according to claim 1, characterized in that, The method described in S105, which involves adjusting feature points on the client side or directly drawing an extended ROI region on the slice, includes: By moving points: By moving specific feature points on the image, based on the movement interaction, these feature points will expand outward spherically by a distance B as the expanded ROI region; Manually delineate the region: By manually drawing a frame, you can delineate a new region of interest on the image. The system will automatically expand the ROI based on the manually delineated position and range. The expanded ROI is the delineated region, and the region is expanded by a thickness C above and below the slice in the vertical direction.

5. A medical image remote transmission device based on an interactive region of interest, characterized in that, include: The module for determining the initial ROI region involves the server determining the initial ROI region based on the segmentation results and feature points. The mask generation module generates an original sampling rate mask based on the initial ROI region and a low sampling rate mask for the entire ROI. The mask refers to the binary volume data that completely overlaps with the original volume data in spatial location. 1 represents that rendering is required and 0 represents that rendering is not required. The data transmission module, based on mask data transmission, includes the following steps: The server generates a raw sampling rate data stream based on the raw sampling rate mask, and extracts the data within the ROI as a voxel data stream. Send the original sampling rate data stream, voxel data stream, downsampled image data, ROI low sampling rate mask, and original sampling rate mask to the client; All data transmissions are compressed, and since the mask data is a 01-byte stream, an extremely high compression ratio can be achieved. The data usage module allows clients to use data in the following two ways: The client restores the complete original image data: interpolates the downsampled image data to restore the voxel interval of the original data and generates new image data V; based on the original sampling rate mask, it takes values ​​from the voxel data stream to replace the corresponding voxels in the image data V; The client simultaneously displays the original sampling rate image data and downsampled image data of the ROI region in space, and performs a mixed display based on multi-volume data; the original sampling rate image data is formed by filling the voxel stream according to the original sampling rate mask; the mixed display supports slice display and volume rendering; The ROI region extension module allows users to extend the ROI region on the client side by adjusting feature points or by directly drawing it on the slice. The sending and receiving module sends the newly generated extended ROI region description information to the server. The server extracts the original sampling rate image based on the received extended ROI region and returns it to the client. The rendering module renders the received raw sampling rate image and modifies the raw sampling rate mask and the ROI low sampling rate mask. The original sampling rate mask is merged and the ROI region is expanded, which means voxel replacement is performed on the binarized mask; The low sampling rate ROI mask is subtracted from the extended ROI region, which is equivalent to voxel replacement of the binarized mask; The image rendering module completes the rendering of images at two sampling rates on the client side using a mask.

Images