Medical technology interaction intelligent funnel breast and chicken breast comprehensive diagnosis and treatment system

Abstract

The invention relates to the technical field of sternum orthopedics, in particular to a medical technology interaction intelligent funnel breast and pigeon breast comprehensive diagnosis and treatment system, which comprises a medical client for collecting a target skeleton image from a patient; the processing platform calculates according to the target skeleton image to obtain target skeleton parameters corresponding to the patient, and then generates bone implant design parameters according to the target skeleton parameters; and the bone implant manufacturing end is used for manufacturing a bone implant corresponding to the patient according to the bone implant design parameters. In order to solve the problems that in the prior art, a prefabricated bone implant cannot well meet the orthopedic requirement, and the use effect is poor, a processing platform is introduced in the scheme to be connected with a medical client on the doctor side to collect a target bone image of a patient, and then a plurality of key indexes in the target bone image are extracted; therefore, the design parameters of the bone implant in the orthopedic process are determined, and finally the bone implant in the corresponding shape is manufactured by the bone implant manufacturing end, so that the surgical effect of the implant is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of sternal correction, and in particular to a medical and technical interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system. Background Art

[0002] The human ribs and sternum protect the heart, lungs, esophagus, respiratory tract and mediastinum in the thoracic cavity. Deformities or injuries will cause great harm to the human body. Changes in the shape of the chest wall can affect the normal function of the organs. Funnel chest can cause heart compression, heart displacement, respiratory diseases, and also affect the appearance.

[0003] Pectus excavatum (PE) is a funnel-shaped deformity in which the middle and lower segments of the sternum and adjacent costal cartilages are concave inward, making the lower part of the anterior chest wall funnel-shaped. The deepest point of the concave is usually at the junction of the lower end of the sternum and the xiphoid process, forming a funnel-like deformity.

[0004] Pectus carinatum is also called pigeon chest. Pectus carinatum is the second most common chest deformity, less common than pectus excavatum. Sometimes pectus excavatum and pectus carinatum may appear at the same time.

[0005] Nuss surgery is a minimally invasive surgery for the treatment of pectus excavatum published by American doctor Nuss in 1998. Nuss surgery is to add a thoracic support Nuss orthopedic plate under the guidance of thoracoscopy to lift up the collapsed sternum, and use a fixator to wrap around the rib wire to fix it to support the collapsed sternum, thereby achieving the purpose of PE chest wall correction. Nuss surgery is a more widely used orthopedic plate-supported correction of pectus excavatum surgery in recent years.

[0006] When pectus excavatum and pectus carinatum appear at the same time, that is, pectus excavatum and pectus carinatum complications, the correction will be more complicated, and sometimes sandwich boards, multiple Nuss boards and special-shaped Nuss boards are needed for correction surgery.

[0007] In the prior art, the Nuss orthotic plate is usually a straight long strip correction plate, which is bent by the doctor based on experience and then implanted into the patient's chest. Its structure is relatively simple and fixed, and it can usually only be used to support chest collapse in one direction. When the chest wall deformity is relatively complex (such as funnel shape and pigeon chest complications), and the body shape of patients of different ages is greatly different from the standard body shape, it cannot meet the correction needs well. In this regard, existing medical device manufacturers are also trying to make a series of bone implants of different shapes to meet the demand, but the actual application still needs to vary according to the "suffering" to achieve the ideal correction effect. Summary of the invention

[0008] In view of the above-mentioned problems existing in the prior art, a medical and technical interactive intelligent comprehensive diagnosis and treatment system for pectus excavatum and pectus carinatum is provided.

[0009] The specific technical solutions are as follows:

[0010] A medical and technical interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system, comprising:

[0011] A medical client, wherein the medical client collects a target bone image of a patient;

[0012] The target bone images include target bones corresponding to pectus carinatum, pectus excavatum and complications;

[0013] A processing platform, the processing platform is connected to the medical client, the processing platform calculates target bone parameters corresponding to the patient according to the target bone image, and then generates bone implant design parameters according to the target bone parameters;

[0014] A bone implant manufacturing end is connected to the processing platform, and the bone implant manufacturing end manufactures a bone implant corresponding to the patient according to the bone implant design parameters.

[0015] In another aspect, the processing platform comprises:

[0016] An image modeling module, wherein the image modeling module performs three-dimensional reconstruction according to the target bone image to obtain a three-dimensional modeling of the target bone;

[0017] A parameter calculation module, the parameter calculation module is connected to the image modeling module, the parameter calculation module extracts a plurality of pairs of calculation reference points for the three-dimensional modeling of the target bone, and then calculates the target bone parameters respectively according to the calculation reference points;

[0018] The target bone parameters include: sternal asymmetry index, sternal correction index, sternal rotation angle, chest concavity length, chest concavity width, and chest concavity depth;

[0019] A bone implant design module is connected to the parameter calculation module, and the bone implant design module generates the bone implant design parameters according to the target bone parameters.

[0020] On the other hand, the image modeling module includes:

[0021] A sternum segmentation module, wherein the sternum segmentation module extracts sternum regions from the target bone image;

[0022] A first registration module, the first registration module is connected to the sternum segmentation module, and the first registration module registers different target bone images in a reconstructed coordinate system to form a registered image group;

[0023] A reconstruction module is connected to the registration module, and the reconstruction module reconstructs the target bone three-dimensional modeling according to the registration image group.

[0024] On the other hand, the image reconstruction module also includes:

[0025] A tissue region labeling module, wherein the tissue region labeling module labels the target bone image to obtain a tissue avoidance region;

[0026] The bone implant design module further includes an avoidance module, which compares the bone implant design parameters generated each time with the tissue avoidance area, and regenerates the bone implant design parameters when the bone implant design parameters conflict with the tissue avoidance area.

[0027] On the other hand, the tissue area labeling module includes:

[0028] A tissue region recognition module, wherein the tissue region recognition module extracts tissue region edges from the target bone image;

[0029] A second registration module, the second registration module is connected to the tissue region identification module, and the second registration module registers the edge of the tissue region in the reconstructed coordinate system to form an edge point set;

[0030] A boundary determination module is connected to the second registration module, and the boundary determination module generates the tissue avoidance area according to the edge point set.

[0031] On the other hand, the tissue region recognition module and the sternum segmentation module both use segmentation models to process the target bone image;

[0032] The medical technology interactive intelligent pectus excavatum carinatum comprehensive diagnosis and treatment system also includes a model training module, which includes:

[0033] A data attribute extraction module, which extracts data attributes from the input target bone image and generates a data fingerprint;

[0034] A pre-training module, the pre-training module is connected to the data attribute extraction module, the pre-training module optimizes a plurality of segmentation networks according to the data fingerprint to obtain an optimized network, and then trains the optimized network;

[0035] A model output module, wherein the model output module is connected to the pre-training module, and the model output module verifies the optimized network and outputs the optimal optimized network as the segmentation model.

[0036] On the other hand, the parameter calculation module includes:

[0037] A first processing module, wherein the first processing module extracts the chest walls on both sides perpendicular to the horizontal line of the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone, and generates the asymmetry index according to the chest walls on both sides;

[0038] A second processing module, which extracts the highest point of the anterior chest wall, the front edge of the thoracic vertebrae, the lowest point of the sternum depression and the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone and generates the correction index;

[0039] A third processing module, which extracts the angle between the sternum plane and the horizontal plane according to the target bone three-dimensional modeling and generates the sternum rotation angle;

[0040] A fourth processing module, wherein the fourth processing module extracts the outer opening of the pectus excavatum depression according to the three-dimensional modeling of the target skeleton and determines the length, width and depth of the chest depression.

[0041] In another aspect, the bone implant design module comprises:

[0042] A difference extraction module, wherein the difference extraction module compares the target bone parameters with reference parameters to extract difference parameters;

[0043] A bone implant prediction module is used to predict the bone implant design parameters according to the difference parameters.

[0044] In another aspect, the bone implant prediction module comprises:

[0045] a generator, wherein the generator predicts intermediate bone implant parameters according to the difference parameters;

[0046] A discriminator is connected to the generator, and the discriminator determines whether the intermediate bone implant parameters meet the optimization conditions, and outputs the intermediate bone implant parameters as the bone implant design parameters when the intermediate bone implant parameters meet the optimization conditions.

[0047] The above technical solution has the following advantages or beneficial effects:

[0048] In view of the problem that the prefabricated bone implants in the existing technology cannot match the orthopedic needs well and have poor use effects, this solution introduces a processing platform to connect to the medical client on the doctor's side to collect the target bone images of the patient, and then extract several key indicators in the target bone images to determine the bone implant design parameters in the orthopedic process. Finally, the bone implant production end produces bone implants of corresponding shapes, thereby improving the surgical effect of the implant. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The embodiments of the present invention will be described more fully with reference to the attached drawings, which are provided for illustration and description only and are not intended to limit the scope of the present invention.

[0050] Figure 1 It is an overall schematic diagram of an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of a processing platform according to an embodiment of the present invention;

[0052] Figure 3 A schematic diagram of an image modeling module according to an embodiment of the present invention;

[0053] Figure 4 This is a schematic diagram of an avoidance module in an embodiment of the present invention;

[0054] Figure 5 This is a schematic diagram of an organization area marking module in an embodiment of the present invention;

[0055] Figure 6 This is a schematic diagram of a model training module in an embodiment of the present invention;

[0056] Figure 7 Schematic diagram of a parameter calculation module in an embodiment of the present invention;

[0057] Figure 8 A schematic diagram of a bone implant design module in an embodiment of the present invention;

[0058] Fig. 9 FIG. 4 is a schematic diagram of a bone implant prediction module in an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0061] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0062] The present invention comprises:

[0063] A medical and technical interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system, such as Figure 1As shown, including:

[0064] Medical client A, medical client A collects target bone images of the patient;

[0065] The target bone images include target bones corresponding to pectus carinatum, pectus excavatum, and comorbidities;

[0066] Processing platform B, processing platform B is connected to medical client A, processing platform B calculates target bone parameters corresponding to the patient according to the target bone image, and then generates bone implant design parameters according to the target bone parameters;

[0067] The bone implant manufacturing end C is connected to the processing platform B, and the bone implant manufacturing end C manufactures the bone implant corresponding to the patient according to the bone implant design parameters.

[0068] Specifically, in response to the problem that prefabricated bone implants in the prior art cannot match the orthopedic needs well and have poor use effects, this solution introduces a processing platform 2 that connects to the medical client 1 on the doctor's side to collect the patient's target bone image, and then extracts several key indicators in the target bone image, thereby determining the bone implant design parameters in the orthopedic process, and finally the bone implant production end 3 produces a bone implant of the corresponding shape, thereby improving the surgical effect of the implant.

[0069] During the implementation process, the above-mentioned medical and technical interactive intelligent funnel chest and carinatum chest comprehensive diagnosis and treatment system is mainly configured in multiple computer devices as a software implementation.

[0070] Among them, medical client A is a client software installed in the doctor's workstation, which is used to obtain the target bone image uploaded by the doctor. The target bone image is usually image data collected by X-ray, MRI and other means in the early stage. In order to facilitate the design of bone implants and subsequent surgical plans, multi-section scanning or three-dimensional imaging is usually performed to collect a series of images related to the patient's bones, especially the bones in the trunk, and the images are stored and uploaded in DICOM image format for processing by processing platform B.

[0071] The target bone images include target bones corresponding to pectus carinatum, pectus excavatum and complications. That is, when a patient has pectus carinatum or pectus excavatum, or complications of pectus carinatum and pectus excavatum, the bone areas that need to be scanned and diagnosed for the patient are the target bones, usually including the sternum, spine and other areas.

[0072] Processing platform B is a server configured with a corresponding processing program, which is used to perform three-dimensional modeling of the input target bone image, obtain a three-dimensional model of the target bone, and identify multiple important points related to sternal deformity as calculation reference points to calculate a series of key indicators as target bone parameters. Finally, the bone implant design is assisted by the algorithm, and the bone implant design parameters are generated and distributed to the bone implant production end C for production.

[0073] The bone implant production terminal C is also a client deployed on the side of the manufacturer or the medical technician department, which is used to receive the bone implant design parameters distributed by the processing platform B, and then produce the bone implant according to the bone implant design parameters, such as receiving the bone implant design parameters in the form of drawings, and then producing the bone implant of the corresponding material through additive manufacturing.

[0074] Bone implants refer to implants used during surgery, especially during the correction of the target bone, such as a bent orthotic plate structure similar to the Nuss orthotic plate, a 3D printed sternum substitute or other equivalent implants.

[0075] In one embodiment, Figure 2 As shown, processing platform B includes:

[0076] Image modeling module 1, image modeling module 1 performs three-dimensional reconstruction according to the target bone image to obtain a three-dimensional modeling of the target bone;

[0077] Parameter calculation module 2, parameter calculation module 2 is connected to image modeling module 1, parameter calculation module extracts multiple pairs of calculation reference points for target bone three-dimensional modeling, and then calculates target bone parameters according to the calculation reference points;

[0078] The target bone parameters include: sternal asymmetry index, sternal correction index, sternal rotation angle, chest concavity length, chest concavity width, and chest concavity depth;

[0079] The bone implant design module 3 is connected to the parameter calculation module 2, and the bone implant design module 3 generates bone implant design parameters according to the target bone parameters.

[0080] Specifically, in order to achieve a better bone implant design process, the inventors first collected and analyzed a large amount of previous bone implant manufacturing case data, and determined that in cases of sternal depression-type deformities, the key points of bone implant manufacturing are the above-mentioned target bone parameters, including asymmetry index, correction index, sternal rotation angle, chest depression length, chest depression width, and chest depression depth.

[0081] Therefore, considering the effectiveness of the manufactured bone implant and the efficiency of model prediction, the selection of the above-mentioned target bone parameters is determined in this embodiment.

[0082] On this basis, firstly, the image modeling module 1 performs three-dimensional reconstruction according to the target bone image to obtain the target bone three-dimensional modeling, then the shape of the target bone three-dimensional modeling is identified by the recognition algorithm, the point selected by the corresponding parameter is determined as the calculation reference point, and then the target bone parameters are calculated, and finally the bone implant design module 3 predicts, generates and iterates the scheme according to the target bone parameters until the implant parameters converge and can meet the correction requirements, and then the bone implant design parameters are generated and output.

[0083] In one embodiment, Figure 3 As shown, the image modeling module 1 includes:

[0084] A sternum segmentation module 11, wherein the sternum segmentation module 11 extracts sternum regions from the target bone image;

[0085] A first registration module 12, the first registration module 12 is connected to the sternum segmentation module 11, and the first registration module 12 registers different target bone images in the reconstructed coordinate system to form a registered image group;

[0086] The reconstruction module 13 is connected to the first registration module 12, and the reconstruction module 13 reconstructs the target bone three-dimensional modeling according to the registration image group.

[0087] Specifically, in order to achieve better three-dimensional reconstruction effects, in this embodiment, the sternum segmentation module 11 first uses a pre-trained segmentation model to extract the sternum area of ​​the target bone image. The segmentation model is a pre-trained artificial intelligence model that can segment and output the required sternum area from the image.

[0088] Then, the first registration module 12 registers each target bone image in the reconstructed coordinate system and remaps the pixel coordinates, especially for the inner and outer edge points of the segmented sternum area and the inner and outer edges of the spine.

[0089] On this basis, the reconstruction module 13 resamples, fits, and connects the edge parts in the registration image group, and then forms triangular surfaces to finally obtain a complete three-dimensional modeling of the target skeleton.

[0090] In one embodiment, Figure 4 As shown, the image modeling module 1 also includes:

[0091] A tissue region labeling module 14, which labels the target bone image to obtain a tissue avoidance region;

[0092] The bone implant design module 2 further includes an avoidance module 21, which compares the bone implant design parameters generated each time with the tissue avoidance area, and regenerates the bone implant design parameters when the bone implant design parameters conflict with the tissue avoidance area.

[0093] Specifically, in order to achieve a better design effect, in the present embodiment, after receiving the target bone image, the tissue area portion contained in the target bone image is also marked to form a tissue avoidance area. During the subsequent bone implant design, an avoidance module 21 is added to verify the scheme generated by the bone implant design module 2. When the bone implant design parameters conflict with the tissue avoidance area, the bone implant design parameters are regenerated to avoid damage to the patient during surgery.

[0094] Furthermore, in the bone implant design module 2, the tissue avoidance area can also be used as a constraint to control the output of the model.

[0095] In one embodiment, Figure 5 As shown, the tissue area marking module 14 includes:

[0096] A tissue region identification module 141, the tissue region identification module 141 extracts tissue region edges from the target bone image;

[0097] A second registration module 142, the second registration module is connected to the tissue region identification module 141, and the second registration module 142 registers the edge of the tissue region in the reconstructed coordinate system to form an edge point set;

[0098] The boundary determination module 143 is connected to the second registration module 142 , and the boundary determination module 143 generates a tissue avoidance region according to the edge point set.

[0099] Specifically, in order to achieve a better tissue region labeling effect, in this embodiment, the tissue region edge is first extracted from the target bone image using a pre-trained segmentation model by the tissue region recognition module 141. Then the second registration module 142 registers the tissue region edge in the reconstructed coordinate system, including unifying the origin of the coordinate points and mapping them to the reconstructed coordinate system, and then forming an edge point set. Finally, the edge point set is described in the form of a function by the boundary determination module 143, and a boundary function is generated as the tissue avoidance region.

[0100] In one embodiment, both the tissue region identification module 141 and the sternum segmentation module 11 use segmentation models to process the target bone image;

[0101] The medical-technical interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system also includes a model training module 4 for training the segmentation model.

[0102] like Figure 6 As shown, the model training module 4 includes:

[0103] A data attribute extraction module 41, the data extraction module 41 extracts data attributes from the input target bone image and generates a data fingerprint;

[0104] A pre-training module 42, the pre-training module 42 is connected to the data attribute extraction module 41, and the pre-training module 42 optimizes the multiple segmentation networks according to the data fingerprint to obtain an optimized network, and then trains the optimized network;

[0105] The model output module 43 is connected to the pre-training module 42, and the model output module 43 verifies the optimized network and outputs the optimal optimized network as a segmentation model.

[0106] Specifically, in order to achieve better model training effects, in this embodiment, the above-mentioned model training module 4 is designed to optimize and train different segmentation networks, thereby achieving effective training and output of target bone images of any segmentation target, thereby improving the efficiency of model training.

[0107] Among them, for the input target bone image, the data extraction module 41 extracts data attributes, including image quality, image modality, image size, voxel size, category ratio, etc., and generates a data fingerprint based on the above attributes.

[0108] Based on the data fingerprint, the pre-training module 42 first optimizes the multiple segmentation networks according to the data fingerprint, including adjusting the size and step size of the variable convolution kernel in the network according to the image size and the image quality of the target bone image actually collected, and adjusting the number of input networks and the number of cascade layers according to the image modality, so as to preliminarily obtain a segmentation network that is more suitable for the target bone image currently generated. Then, the adjusted segmentation network is trained using the sample data corresponding to the target bone image, and finally, the model output module 43 verifies the optimized network and outputs the optimal optimized network as the segmentation model.

[0109] The segmentation network includes multiple typical models, such as U-Net network, CNN network, etc., which usually have multi-head input and image segmentation functions.

[0110] In one embodiment, Figure 7 As shown, the parameter calculation module 2 includes:

[0111] A first processing module 21, the first processing module 21 extracts the chest walls on both sides perpendicular to the horizontal line of the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target skeleton, and generates an asymmetry index according to the chest walls on both sides;

[0112] Specifically, after establishing the three-dimensional modeling of the target bone, the first processing module 21 first identifies the left and right sides of the three-dimensional modeling of the target bone by reconstructing the coordinate system, determines the two side boundaries of the maximum transverse diameter of the thorax, and then extracts the two sides of the chest wall perpendicular to the horizontal line of the maximum transverse diameter of the thorax, and records the anterior diameter of the left chest wall as L, and the anterior diameter of the right chest wall as R, and then calculates the asymmetry index AI = 1-R / L to provide a basis for the subsequent measurement of the degree of chest wall asymmetry.

[0113] The second processing module 22 extracts the highest point of the anterior chest wall, the front edge of the thoracic vertebrae, the lowest point of the sternum depression and the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone and generates a correction index;

[0114] Similarly, in the reconstructed coordinate system, the second processing module 22 first extracts the highest point of the anterior chest wall, the front edge of the thoracic vertebrae, the lowest point of the sternal recess and the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone, and then calculates the difference between the vertical distance from the highest point of the anterior chest wall to the front edge of the thoracic vertebrae and the vertical distance from the lowest point of the sternal recess to the front edge of the thoracic vertebrae and the maximum transverse diameter of the thorax in the same plane as the correction index.

[0115] A third processing module 23, the third processing module 23 extracts the angle between the sternum plane and the horizontal plane according to the target bone three-dimensional modeling and generates a sternum rotation angle;

[0116] Similarly, in the reconstructed coordinate system, the third processing module 23 performs three-dimensional modeling of the target skeleton to extract the sternum plane, which is determined by fitting multiple vertices of the sternum, and then calculates the angle between the sternum plane and the horizontal plane to generate the sternum rotation angle.

[0117] The fourth processing module 24 extracts the outer opening of the pectus excavatum depression according to the target bone three-dimensional modeling and determines the length, width and depth of the chest cavity depression.

[0118] Furthermore, in the reconstructed coordinate system, the fourth processing module 24 performs three-dimensional modeling of the target skeleton to determine the points on the front of the sternum, calculates the curvature of each point according to the points on the front of the sternum, thereby determining the external opening of the funnel chest depression, and then calculates the length, width and depth of the thoracic depression.

[0119] In one embodiment, Figure 8 As shown, the bone implant design module 3 includes:

[0120] A difference extraction module 31, the difference extraction module 31 compares the target bone parameters with the reference parameters to extract the difference parameters that are different;

[0121] The bone implant prediction module 32 predicts the bone implant design parameters according to the difference parameters.

[0122] Specifically, in order to achieve a better bone correction effect, in this embodiment, the target bone parameters are first compared with the reference parameters by the difference extraction module 31 to extract the difference parameters with differences.

[0123] Specifically, during the early data collection process, the patient's physiological parameters are obtained, mainly including chest diameter, trunk length, etc. According to these physiological parameters, the expected corrected three-dimensional chest structure can be generated. Then, the difference extraction module 31 compares the target bone parameters with the reference parameters to obtain the difference parameters with differences. The difference parameters refer to the coordinate offsets of the multiple landmarks on each sternum relative to the corresponding points on the expected corrected structure in the reconstructed coordinate system. These difference parameters are the parameters that need to be corrected by bone implants in the future.

[0124] On this basis, the bone implant prediction module 32 uses an artificial intelligence model to predict the bone implant design parameters according to the difference parameters. Taking Nuss surgery as an example, it includes the optimal placement of the NUSS plate to ensure that it can effectively support the depressed area; the appropriate length to match the patient's chest size; and the necessary curvature to fit the natural curve of the patient's chest to achieve the best correction effect.

[0125] In one embodiment, Fig. 9 As shown, the bone implant prediction module 32 includes:

[0126] A generator 321, the generator 321 predicts intermediate bone implant parameters according to the difference parameters;

[0127] The discriminator 322 is connected to the generator 321, and the discriminator 322 determines whether the intermediate bone implant parameters meet the optimization conditions, and outputs the intermediate bone implant parameters as bone implant design parameters when the optimization conditions are met.

[0128] Specifically, in order to achieve a better bone implant design effect, in this embodiment, a pre-trained adversarial generative network is introduced in the bone implant prediction module 32. Among them, the generator 321 predicts the corresponding bone implant design parameters according to the difference parameters, that is, the sternum points to be corrected as the intermediate bone implant parameters, and then the discriminator 322 determines whether the intermediate bone implant parameters can be used to correct the specific points and whether they meet the optimization conditions, such as weight, total volume, whether they invade the marked tissue area, etc., and outputs the intermediate bone implant parameters as bone implant design parameters when the optimization conditions are met, thereby achieving a better design effect.

[0129] Those skilled in the art will appreciate that various aspects of the present invention, or possible implementations of various aspects, may be specifically implemented as systems, methods, or computer program products. Therefore, various aspects of the present invention, or possible implementations of various aspects, may take the form of complete hardware embodiments, complete software embodiments (including firmware, resident software, etc.), or embodiments of combined software and hardware aspects, all collectively referred to herein as "circuits," "modules," or "systems." In addition, various aspects of the present invention, or possible implementations of various aspects, may take the form of computer program products, which refer to computer instructions stored in a memory.

[0130] The memory may be a computer-readable signal medium or a computer-readable storage medium. Computer-readable storage media include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or apparatuses, or any suitable combination of the foregoing, such as random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable read-only memory (CD-ROM).

[0131] The processor in the computer reads the computer instructions stored in the memory, so that the processor can execute the functional actions specified in each step or the combination of steps in the flowchart; and generate a device for implementing the functional actions specified in each block or the combination of blocks in the block diagram.

[0132] It should be understood that the processor in the computer can be understood as one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components implemented to execute the aforementioned computer instructions.

[0133] Computer instructions can be executed completely on the user's local computer, partially on the user's local computer, as a separate software package, partially on the user's local computer and partially on a remote computer, or completely on a remote computer or server. It should also be noted that in some alternative embodiments, the functions noted in each step in the flow chart or each block in the block diagram may not occur in the order noted in the figure. For example, depending on the functions involved, two steps or two blocks shown in succession may actually be executed roughly simultaneously, or these blocks may sometimes be executed in reverse order.

[0134] The above are only preferred embodiments of the present invention, and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A medical and technical interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system, characterized in that: include: A medical client, wherein the medical client collects a target bone image of a patient; The target bone images include target bones corresponding to pectus carinatum, pectus excavatum and complications; A processing platform, the processing platform is connected to the medical client, the processing platform calculates target bone parameters corresponding to the patient according to the target bone image, and then generates bone implant design parameters according to the target bone parameters; A bone implant manufacturing end is connected to the processing platform, and the bone implant manufacturing end manufactures a bone implant corresponding to the patient according to the bone implant design parameters.

2. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 1 is characterized in that: The processing platform comprises: An image modeling module, wherein the image modeling module performs three-dimensional reconstruction according to the target bone image to obtain a three-dimensional modeling of the target bone; A parameter calculation module, the parameter calculation module is connected to the image modeling module, the parameter calculation module extracts a plurality of pairs of calculation reference points for the three-dimensional modeling of the target bone, and then calculates the target bone parameters respectively according to the calculation reference points; The target bone parameters include: sternal asymmetry index, sternal correction index, sternal rotation angle, chest concavity length, chest concavity width, and chest concavity depth; A bone implant design module is connected to the parameter calculation module, and the bone implant design module generates the bone implant design parameters according to the target bone parameters.

3. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 2 is characterized in that: The image modeling module comprises: A sternum segmentation module, wherein the sternum segmentation module extracts sternum regions from the target bone image; A first registration module, the first registration module is connected to the sternum segmentation module, and the first registration module registers different target bone images in a reconstructed coordinate system to form a registered image group; A reconstruction module, wherein the reconstruction module is connected to the first registration module, and the reconstruction module reconstructs the target bone three-dimensional modeling according to the registration image group.

4. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 3 is characterized in that: The image modeling module also includes: A tissue region labeling module, wherein the tissue region labeling module labels the target bone image to obtain a tissue avoidance region; The bone implant design module further includes an avoidance module, which compares the bone implant design parameters generated each time with the tissue avoidance area, and regenerates the bone implant design parameters when the bone implant design parameters conflict with the tissue avoidance area.

5. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 4 is characterized in that: The tissue area marking module includes: A tissue region recognition module, wherein the tissue region recognition module extracts tissue region edges from the target bone image; A second registration module, the second registration module is connected to the tissue region identification module, and the second registration module registers the edge of the tissue region in the reconstructed coordinate system to form an edge point set; A boundary determination module is connected to the second registration module, and the boundary determination module generates the tissue avoidance area according to the edge point set.

6. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 5 is characterized in that: The tissue region recognition module and the sternum segmentation module both use segmentation models to process the target bone image; The medical technology interactive intelligent pectus excavatum carinatum comprehensive diagnosis and treatment system also includes a model training module, which includes: A data attribute extraction module, which extracts data attributes from the input target bone image and generates a data fingerprint; A pre-training module, the pre-training module is connected to the data attribute extraction module, the pre-training module optimizes a plurality of segmentation networks according to the data fingerprint to obtain an optimized network, and then trains the optimized network; A model output module, wherein the model output module is connected to the pre-training module, and the model output module verifies the optimized network and outputs the optimal optimized network as the segmentation model.

7. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 2 is characterized in that: The parameter calculation module comprises: A first processing module, wherein the first processing module extracts the chest walls on both sides perpendicular to the horizontal line of the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone, and generates the asymmetry index according to the chest walls on both sides; A second processing module, which extracts the highest point of the anterior chest wall, the front edge of the thoracic vertebrae, the lowest point of the sternum depression and the maximum transverse diameter of the thorax according to the three-dimensional modeling of the target bone and generates the correction index; A third processing module, which extracts the angle between the sternum plane and the horizontal plane according to the target bone three-dimensional modeling and generates the sternum rotation angle; A fourth processing module, wherein the fourth processing module extracts the outer opening of the pectus excavatum depression according to the three-dimensional modeling of the target skeleton and determines the length, width and depth of the chest depression.

8. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 1 is characterized in that: The bone implant design module includes: A difference extraction module, wherein the difference extraction module compares the target bone parameters with reference parameters to extract difference parameters; A bone implant prediction module is used to predict the bone implant design parameters according to the difference parameters.

9. The medical technology interactive intelligent pectus excavatum and pectus carinatum comprehensive diagnosis and treatment system according to claim 8 is characterized in that: The bone implant prediction module comprises: a generator, wherein the generator predicts intermediate bone implant parameters according to the difference parameters; A discriminator is connected to the generator, and the discriminator determines whether the intermediate bone implant parameters meet the optimization conditions, and outputs the intermediate bone implant parameters as the bone implant design parameters when the intermediate bone implant parameters meet the optimization conditions.

Images