Method, device and medium for cut margin path generation for body surface tumors

Abstract

The application provides a method and device for generating a tumor margin path on a body surface, equipment and a medium, which comprehensively considers 3D structured light camera optical data of the body surface tumor and three-dimensional high-frequency ultrasonic data of the body surface tumor obtained by a three-dimensional high-frequency ultrasound instrument. The 3D structured light camera optical data of the body surface tumor can obtain clear and accurate three-dimensional directional contour boundary information of the surface of the body surface tumor exposed to the outside, and the three-dimensional high-frequency ultrasonic data of the body surface tumor can obtain three-dimensional boundary information of the overall contour of the body surface tumor including the depth direction. The three-dimensional high-frequency ultrasonic data of the body surface tumor and the 3D structured light camera optical data of the body surface tumor are registered and fused, and based on this, target three-dimensional data at any position of the body surface tumor is obtained, which can accurately reflect three-dimensional information of the surface boundary and the depth direction boundary of the body surface tumor, and the finally generated three-dimensional margin path can greatly improve the accuracy of the margin path of the body surface tumor.

Description

Technical Field

[0001] This invention mainly relates to the field of computer-aided preoperative analysis technology, and in particular to a method, device, equipment and medium for generating surgical margin paths for tumors on the body surface. Background Technology

[0002] Superficial tumors are among the most common cancers in humans, arising from malignant proliferative diseases of cells in the skin and subcutaneous tissues. Malignant tumors can proliferate continuously and metastasize to distant sites, seriously threatening life and health. Surgical resection is one of the effective treatment methods. During surgery, surgeons generally design the resection margins based on their experience in judging the shape and size before performing the surgical resection.

[0003] In clinical practice, the design of surgical margins for superficial tumors is generally based on the physician's own experience, either by sketching the margins or by using appropriate methods to automatically generate them. Physicians' use of their own experience in margin design relies heavily on their subjective judgment, resulting in poor objectivity. Furthermore, existing automatic generation methods do not address the tumor margin path along the depth of the lesion, nor do they consider the impact of cross-plane travel on the margin path when the lesion surface is curved.

[0004] In summary, existing methods cannot guarantee the accuracy of incisions in three dimensions. Incomplete tumor removal increases the risk of recurrence and malignant transformation, while excessive removal increases patient trauma and affects prognosis. Furthermore, the tumor is surrounded by a rich blood supply; inaccurate incision planning may lead to the surgeon severing blood vessels during surgery, affecting the surgical field and increasing the risk of tumor metastasis. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention proposes a method, apparatus, device, and medium for generating surgical margin paths for tumors on the body surface.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] On one hand, the present invention provides a method for generating surgical margin paths for tumors on the body surface, comprising:

[0008] Input the optical data of the 3D structured light camera of the tumor on the body surface, and obtain the three-dimensional optical image data of the exposed surface contour of the tumor.

[0009] In three-dimensional optical image data, the exposed tumor on the outer surface of the body is discretized into a dense lattice of points as the first feature point cloud.

[0010] Input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and obtain the three-dimensional acoustic image data of the tumor on the body surface;

[0011] Reconstructing the surface contour of a body surface tumor from three-dimensional acoustic image data based on the three-dimensional acoustic image data of the body surface tumor;

[0012] A dense lattice of points is discretized on the surface contour of the tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0013] The first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire body surface tumor that reflects the surface boundary and depth direction boundary of the tumor.

[0014] The three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface is segmented and extracted to obtain the initial three-dimensional cutting edge path;

[0015] The final three-dimensional cutting edge path is generated based on the initial three-dimensional cutting edge path.

[0016] On the other hand, the present invention provides a device for generating surgical margin paths for tumors on the body surface, comprising:

[0017] The first module is used to input the optical data of the 3D structured light camera of the tumor on the body surface and obtain the three-dimensional optical image data of the outline of the tumor exposed on the outer surface.

[0018] The second module is used to discretize a dense lattice of points on the outer surface contour of the exposed tumor in the three-dimensional optical image data as the first feature point cloud.

[0019] The third module is used to input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and to obtain the three-dimensional acoustic image data of the tumor on the body surface.

[0020] The fourth module is used to reconstruct the surface contour of the tumor in the three-dimensional acoustic image data based on the three-dimensional acoustic image data of the tumor surface.

[0021] The fifth module is used to discretize a dense lattice of points on the surface contour of a tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0022] The sixth module is used to register and fuse the first feature point cloud and the second feature point cloud to obtain a three-dimensional image of the entire tumor surface that reflects the surface boundary and depth direction boundary of the tumor.

[0023] The seventh module is used to segment and extract the three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface to obtain the initial three-dimensional cutting edge path;

[0024] The eighth module is used to generate the final three-dimensional cutting edge path based on the initial three-dimensional cutting edge path.

[0025] On the other hand, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0026] Input the optical data of the 3D structured light camera of the tumor on the body surface, and acquire the three-dimensional optical image data of the outline of the tumor exposed on the outer surface.

[0027] In three-dimensional optical image data, the exposed tumor on the outer surface of the body is discretized into a dense lattice of points as the first feature point cloud.

[0028] Input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and obtain the three-dimensional acoustic image data of the tumor on the body surface;

[0029] Reconstructing the surface contour of a body surface tumor from three-dimensional acoustic image data based on the three-dimensional acoustic image data of the body surface tumor;

[0030] A dense lattice of points is discretized on the surface contour of the tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0031] The first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire body surface tumor that reflects the surface boundary and depth direction boundary of the tumor.

[0032] The three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface is segmented and extracted to obtain the initial three-dimensional cutting edge path;

[0033] The final three-dimensional cutting edge path is generated based on the initial three-dimensional cutting edge path.

[0034] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, performs the following steps:

[0035] Input the optical data of the 3D structured light camera of the tumor on the body surface, and acquire the three-dimensional optical image data of the outline of the tumor exposed on the outer surface.

[0036] In three-dimensional optical image data, the exposed tumor on the outer surface of the body is discretized into a dense lattice of points as the first feature point cloud.

[0037] Input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and obtain the three-dimensional acoustic image data of the tumor on the body surface;

[0038] Reconstructing the surface contour of a body surface tumor from three-dimensional acoustic image data based on the three-dimensional acoustic image data of the body surface tumor;

[0039] A dense lattice of points is discretized on the surface contour of the tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0040] The first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire body surface tumor that reflects the surface boundary and depth direction boundary of the tumor.

[0041] The three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface is segmented and extracted to obtain the initial three-dimensional cutting edge path;

[0042] The final three-dimensional cutting edge path is generated based on the initial three-dimensional cutting edge path.

[0043] Compared with the prior art, the technical effects of the present invention are as follows:

[0044] This invention comprehensively considers the optical data of a 3D structured light camera for tumors on the body surface and the 3D high-frequency ultrasound data of a 3D high-frequency ultrasound instrument for tumors on the body surface. The optical data of the 3D structured light camera for tumors on the body surface can obtain clear and accurate 3D contour boundary information of the exposed surface of the tumor, while the 3D high-frequency ultrasound data can obtain the 3D boundary information of the tumor contour, including the depth direction.

[0045] This invention registers and fuses three-dimensional high-frequency ultrasound data of tumors on the body surface with optical data of tumors in the 3D structured light camera. Based on this, the target three-dimensional data at any location of the tumor on the body surface is obtained, which can accurately reflect the three-dimensional information of the surface boundary and depth direction boundary of the tumor. Based on this, the final three-dimensional cutting edge path is generated, which can greatly improve the accuracy of the cutting edge path of tumors on the body surface. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0047] Figure 1 This is a flowchart of a method for generating surgical margin paths for tumors on the body surface, provided in one embodiment;

[0048] Figure 2 This is a schematic diagram of the three-dimensional resection boundary of a tumor on the body surface in one embodiment. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0050] Reference Figure 1 One embodiment provides a method for generating surgical margin paths for tumors on the body surface, comprising:

[0051] Input the optical data of the 3D structured light camera of the tumor on the body surface, and acquire the three-dimensional optical image data of the outline of the tumor exposed on the outer surface.

[0052] In three-dimensional optical image data, the exposed tumor on the outer surface of the body is discretized into a dense lattice of points as the first feature point cloud.

[0053] Input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and obtain the three-dimensional acoustic image data of the tumor on the body surface;

[0054] Reconstructing the surface contour of a body surface tumor from three-dimensional acoustic image data based on the three-dimensional acoustic image data of the body surface tumor;

[0055] A dense lattice of points is discretized on the surface contour of the tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0056] The first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire body surface tumor that reflects the surface boundary and depth direction boundary of the tumor.

[0057] The three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface is segmented and extracted to obtain the initial three-dimensional cutting edge path;

[0058] The final three-dimensional cutting edge path is generated based on the initial three-dimensional cutting edge path.

[0059] Optical data from a 3D structured light camera can obtain clear and accurate three-dimensional contour boundary information of the tumor surface, while three-dimensional high-frequency ultrasound data of the tumor surface can obtain three-dimensional boundary information of the tumor contour. The three-dimensional image of the entire tumor surface obtained by this invention, based on optical data from a 3D structured light camera and three-dimensional high-frequency ultrasound data of the tumor surface, accurately reflects the three-dimensional information of the tumor surface boundary and depth direction boundary, providing a more realistic and reliable basis for subsequent accurate segmentation and extraction of the three-dimensional contour boundary of the tumor surface.

[0060] Furthermore, the first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire tumor surface reflecting the surface boundary and depth direction boundary of the tumor, including:

[0061] The acquired second feature point cloud is processed using a 3D local feature description algorithm to extract representative point cloud features. Then, a fast point feature value histogram is used to obtain the initial correspondence point set between the second and first feature point clouds. A random sampling consensus algorithm is then used to remove mismatched points between the initial second and first feature point clouds, resulting in a corrected correspondence point set. This completes a coarse matching of the second and first feature point clouds. Finally, based on the corrected correspondence point set, a nearest-point iteration algorithm is used for fine registration of the second and first feature point clouds, yielding a three-dimensional image of the entire tumor surface reflecting its surface and depth boundaries. Thus, this method achieves high-precision registration of the first and second feature point clouds, enabling the final three-dimensional image of the entire tumor surface to more realistically and accurately reflect its surface and depth boundaries.

[0062] The three-dimensional image of the entire body surface tumor is input into the trained path generation model. The trained path generation model is used to classify and segment the three-dimensional image of the entire body surface tumor, extract the three-dimensional contour boundary of the body surface tumor, and obtain the prediction result of the body surface tumor type and the initial three-dimensional cutting edge path.

[0063] One embodiment proposes the path generation model, which is obtained through the following method:

[0064] Construct a large amount of training data and the label information of the training data, wherein the training data is a three-dimensional image of the entire body surface tumor obtained after registration and fusion, and the label information of the training data is the type of body surface tumor in the three-dimensional image of the entire body surface tumor obtained after registration and fusion.

[0065] The training data and its label information are used as input to the deep learning network model, and the tumor type and the three-dimensional contour boundary of the tumor are used as output. The model is trained iteratively to obtain a well-trained path generation model.

[0066] It is understandable that when training a path generation model, the training data can be randomly divided into a training set and a validation set in an 8:2 ratio. The training set data is used to build the model, and the validation set data is used for further testing and adjustment of the model. A path generation model for predicting tumor types and resection margins is established based on a convolutional neural network. This model can identify tumor types and resection margins based on the entire tumor lesion image obtained after registration and fusion.

[0067] One embodiment proposes generating a final three-dimensional resection path based on the initial three-dimensional resection path and the type of tumor on the body surface, including:

[0068] When the tumor type on the body surface is benign, the initial three-dimensional cutting edge path is smoothed in three dimensions to generate a fusiform cutting edge path as the final three-dimensional cutting edge path.

[0069] In the case of malignant tumors on the body surface, the initial cutting edge path is enlarged equidistantly in three dimensions to generate the final three-dimensional cutting edge path.

[0070] Figure 2 This is a schematic diagram of the three-dimensional resection boundary of a tumor on the body surface in one embodiment. (Refer to...) Figure 2 The central dark region is defined by the initial three-dimensional cutting edge path. Expanding this central dark region yields the region defined by the final three-dimensional cutting edge path. Let the coordinates of a point 'a' on the initial three-dimensional cutting edge path of the tumor surface, i.e., on the boundary of the central dark region, be (x...). a ,y a ,z a If the boundary point a(x) of the dark region in the middle of the initial three-dimensional cutting edge path is then determined, then... a ,y a ,z a The point corresponding to the final 3D cutting edge path generated after expansion is point b, and the coordinates of point b are (x...). b ,y b ,z b The length of the arc along the skin surface between points a and b is the expansion distance l. b That is, l b =[(x b -x a ) 2 +(y b -y a ) 2 +(z b -z a ) 2 ] 1 / 2 .

[0071] In one embodiment, a method for generating a final three-dimensional resection path based on the initial three-dimensional resection path and the type of tumor on the body surface is proposed, including:

[0072] When the tumor type on the body surface is benign, the final three-dimensional cutting edge path is generated by setting the outward cutting distance from the initial three-dimensional cutting edge path to 1mm-5mm.

[0073] When the tumor type on the body surface is malignant, the final three-dimensional cutting edge path is generated by setting the outward cutting distance from the initial three-dimensional cutting edge path to 4mm-30mm.

[0074] In one embodiment, when the tumor type on the body surface is a malignant tumor and is basal cell carcinoma, the final three-dimensional cutting edge path is generated by setting the outward cutting distance based on the initial three-dimensional cutting edge path to 4 mm.

[0075] When the tumor type on the body surface is malignant and is melanoma, if the tumor thickness is less than 1 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 10 mm; if the tumor thickness is between 1 and 2 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 10 to 20 mm; if the tumor thickness is greater than 2 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 20 mm; if the tumor thickness is greater than 4 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 30 mm.

[0076] When the tumor type on the body surface is malignant and is primary low-risk squamous cell carcinoma of the skin, if the diameter of the tumor is less than 20 mm, the outward expansion distance based on the initial three-dimensional cutting edge path is set to 4 mm to generate the final three-dimensional cutting edge path; if the diameter of the tumor on the body surface is greater than or equal to 20 mm, the outward expansion distance based on the initial three-dimensional cutting edge path is set to 6 mm to generate the final three-dimensional cutting edge path.

[0077] When the tumor on the body surface is a malignant tumor and is a primary high-risk squamous cell carcinoma of the skin, if the diameter of the tumor is less than 10 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of at least 4 mm; if the diameter of the tumor on the body surface is between 10 and 19 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of at least 6 mm; if the diameter of the tumor on the body surface is greater than or equal to 20 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of at least 9 mm.

[0078] To better display the final three-dimensional resection path, this invention may further include displaying the fused three-dimensional image of the entire tumor surface and the final three-dimensional resection path on a monitor. Real-time display on the monitor provides visual operation prompts, reducing the influence of human factors and making the process more intuitive and accurate.

[0079] One embodiment provides a device for generating surgical margin paths for tumors on the body surface, comprising:

[0080] The first module is used to input the optical data of the 3D structured light camera of the tumor on the body surface and obtain the three-dimensional optical image data of the outline of the tumor exposed on the outer surface.

[0081] The second module is used to discretize a dense lattice of points on the outer surface contour of the exposed tumor in the three-dimensional optical image data as the first feature point cloud.

[0082] The third module is used to input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and to obtain the three-dimensional acoustic image data of the tumor on the body surface.

[0083] The fourth module is used to reconstruct the surface contour of the tumor in the three-dimensional acoustic image data based on the three-dimensional acoustic image data of the tumor surface.

[0084] The fifth module is used to discretize a dense lattice of points on the surface contour of a tumor in the three-dimensional acoustic image data as a second feature point cloud.

[0085] The sixth module is used to register and fuse the first feature point cloud and the second feature point cloud to obtain a three-dimensional image of the entire tumor surface that reflects the surface boundary and depth direction boundary of the tumor.

[0086] The seventh module is used to segment and extract the three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface to obtain the initial three-dimensional cutting edge path;

[0087] The eighth module is used to generate the final three-dimensional cutting edge path based on the initial three-dimensional cutting edge path.

[0088] The implementation methods of the above modules and the construction of the model can all adopt the methods described in any of the foregoing embodiments, and will not be repeated here.

[0089] This invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the method for generating surgical margins for tumor surfaces provided in any of the above embodiments. The computer device can be a server. The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device stores sample data. The network interface of the computer device is used for communication with external terminals via a network connection.

[0090] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for generating surgical margins for tumors on the body surface provided in any of the above embodiments.

[0091] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0092] Matters not covered in this invention are common knowledge.

[0093] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0094] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for generating surgical margin paths for tumors on the body surface, characterized in that, include: Input the optical data of the 3D structured light camera of the tumor on the body surface, and obtain the three-dimensional optical image data of the surface contour of the tumor exposed on the body surface; A dense lattice of points is discretized on the exposed surface contour of the tumor in the three-dimensional optical image data as the first feature point cloud. Input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and obtain the three-dimensional acoustic image data of the tumor on the body surface; Reconstructing the surface contour of a body surface tumor from three-dimensional acoustic image data based on the three-dimensional acoustic image data of the body surface tumor; A dense lattice of points is discretized on the surface contour of the tumor in the three-dimensional acoustic image data as a second feature point cloud. The first feature point cloud and the second feature point cloud are registered and fused to obtain a three-dimensional image of the entire body surface tumor that reflects the surface boundary and depth direction boundary of the tumor. The three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface is segmented and extracted to obtain the initial three-dimensional cutting edge path; The process of generating the final three-dimensional cutting edge path based on the initial three-dimensional cutting edge path includes: When the tumor type on the body surface is benign, the final three-dimensional cutting edge path is generated by setting the outward cutting distance from the initial three-dimensional cutting edge path to 1mm~5mm. When the tumor on the body surface is malignant, the final three-dimensional resection path is generated by setting the outward expansion distance from the initial three-dimensional resection path to 4mm~30mm, where: When the tumor type on the body surface is a malignant tumor and is basal cell carcinoma, the final three-dimensional cutting edge path is generated by setting the outward cutting distance based on the initial three-dimensional cutting edge path to 4mm. When the tumor type on the body surface is malignant and is melanoma, if the tumor thickness is less than 1 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 10 mm; if the tumor thickness is between 1 and 2 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 10 to 20 mm; if the tumor thickness is greater than 2 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 20 mm; if the tumor thickness is greater than 4 mm, the final three-dimensional cutting margin path is generated by expanding the initial three-dimensional cutting margin path outward by a distance of 30 mm. When the tumor type on the body surface is malignant and is primary low-risk squamous cell carcinoma of the skin, if the diameter of the tumor is less than 20 mm, the outward expansion distance based on the initial three-dimensional cutting edge path is set to 4 mm to generate the final three-dimensional cutting edge path; if the diameter of the tumor on the body surface is greater than or equal to 20 mm, the outward expansion distance based on the initial three-dimensional cutting edge path is set to 6 mm to generate the final three-dimensional cutting edge path. When the tumor on the body surface is a malignant tumor and is primary high-risk squamous cell carcinoma of the skin, if the tumor diameter is less than 10 mm, the final three-dimensional cutting margin is generated by expanding the initial three-dimensional cutting margin path outward by a distance greater than 4 mm; if the tumor diameter is between 10 and 19 mm, the final three-dimensional cutting margin is generated by expanding the initial three-dimensional cutting margin path outward by a distance greater than 6 mm; if the tumor diameter is greater than or equal to 20 mm, the final three-dimensional cutting margin is generated by expanding the initial three-dimensional cutting margin path outward by a distance greater than 9 mm.

2. The method for generating surgical margin paths for tumors on the body surface according to claim 1, characterized in that, The three-dimensional image of the entire body surface tumor is input into the trained path generation model. The trained path generation model is used to classify and segment the three-dimensional image of the entire body surface tumor, extract the three-dimensional contour boundary of the body surface tumor, and obtain the prediction result of the body surface tumor type and the initial three-dimensional cutting edge path.

3. The method for generating surgical margin paths for tumors on the body surface according to claim 1, characterized in that, The path generation model is obtained in the following way: Construct a large amount of training data and the label information of the training data, wherein the training data is a three-dimensional image of the entire body surface tumor obtained after registration and fusion, and the label information of the training data is the type of body surface tumor in the three-dimensional image of the entire body surface tumor obtained after registration and fusion. The training data and its label information are used as input to the deep learning network model, and the tumor type and the three-dimensional contour boundary of the tumor are used as output. The model is trained iteratively to obtain a well-trained path generation model.

4. The method for generating surgical margin paths for tumors on the body surface according to any one of claims 1 to 3, characterized in that, It also includes displaying the final 3D cutting path on the monitor.

5. A device for generating surgical margin paths for tumors on the body surface, used to implement the method for generating surgical margin paths for tumors on the body surface as described in any one of claims 1 to 3, characterized in that, include: The first module is used to input the optical data of the 3D structured light camera of the tumor on the body surface and obtain the three-dimensional optical image data of the surface contour of the tumor exposed on the body surface. The second module is used to discretize a dense lattice of points on the exposed surface contour of the tumor in the three-dimensional optical image data as the first feature point cloud. The third module is used to input the three-dimensional high-frequency ultrasound data of the tumor on the body surface from the three-dimensional high-frequency ultrasound instrument, and to obtain the three-dimensional acoustic image data of the tumor on the body surface. The fourth module is used to reconstruct the surface contour of the tumor in the three-dimensional acoustic image data based on the three-dimensional acoustic image data of the tumor surface. The fifth module is used to discretize a dense lattice of points on the surface contour of a tumor in the three-dimensional acoustic image data as a second feature point cloud. The sixth module is used to register and fuse the first feature point cloud and the second feature point cloud to obtain a three-dimensional image of the entire tumor surface that reflects the surface boundary and depth direction boundary of the tumor. The seventh module is used to segment and extract the three-dimensional contour boundary of the tumor in the entire three-dimensional image of the tumor surface to obtain the initial three-dimensional cutting edge path; The eighth module is used to generate the final three-dimensional cutting edge path based on the initial three-dimensional cutting edge path.

6. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes a computer program, it implements the steps of the method for generating surgical margin paths for tumors on the body surface as described in claim 1.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the steps of the method for generating surgical margin paths for tumors on the body surface as described in claim 1.

Images