Fracture reduction dynamic planning system based on three-dimensional reconstruction technology

Abstract

The application relates to the field of computer-assisted surgery, in particular to a fracture reduction dynamic planning system based on three-dimensional reconstruction technology, which comprises the following steps: acquiring a reference position and a reference path which are closest to a target position and have the most similar bone posture on all candidate paths outside a target path; recording the similarity of the constraint element posture at the position before the target position on the target path and the constraint element posture at the position before the reference position on the reference path as a path switching index of the target position; recording the target position with the path switching index greater than a preset threshold as a path switching point; and acquiring an optimal path and a backup reduction path based on the path switching point. The application can establish a dynamic switching mechanism based on multiple backup reduction paths when a reduction error occurs in the fracture reduction process, can realize error tolerance through path switching index optimization, and can avoid the situation that fracture reduction cannot be completed as expected when a single reduction path has a reduction error.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of computer-assisted surgery, in particular to a fracture reduction dynamic planning system based on three-dimensional reconstruction technology. BACKGROUND

[0002] The fracture reduction process is usually completed based on surgical navigation technology, which relies on the preoperative fracture reduction path. The existing fracture reduction path is usually completed based on computer-aided planning technology, such as dynamically planning the fracture reduction path by reconstructing the musculoskeletal model. The prior art in dynamically planning the fracture reduction path, such as the patent with publication number CN116370073A, a lower limb bone high-energy injury reduction navigation method, focuses on obtaining an optimal path through optimization simulation, so that the fracture reduction process is optimal in some aspects.

[0003] However, in the fracture reduction process, due to the complexity of human tissues, the optimal fracture reduction path designed preoperatively may have a reduction error when reduction is performed according to the optimal fracture reduction path. The reduction error refers to the fact that the position and posture of the sub-bone do not change according to the fracture reduction path during the sub-bone reduction process, but deviate from the designed fracture reduction path; this leads to the fact that the subsequent fracture reduction process needs to rely on the experience of doctors to complete, or the fracture reduction result deviates significantly from the expectation, or even when the reduction is continued according to the preoperative fracture reduction path, it will cause significant secondary damage to the patient;

[0004] For example, the intelligent reduction robot system in pelvic fracture has an average reduction error of 3.41mm, and the postoperative residual displacement reaches 4.61mm, indicating that even if automatic technology is used when only one reduction path is used, there is still an unavoidable deviation; for another example, Lei Jingtao, Wang Yang, Cheng Liya of Shanghai University and Hu Lei, Wang Tianmiao of Beijing University of Aeronautics and Astronautics published a paper entitled "Reduction Robot Safety Strategy Based on Reduction Path Envelope Error and Improved Artificial Potential Field Method" in the January issue of the Journal of Mechanical Engineering in 2020, which records that in long bone fracture reduction, if the bone block deviates more than the envelope error (usually set to 2mm), it needs to rely on the experience of doctors to adjust.

[0005] In summary, the prior art only uses one reduction path, and when a reduction error occurs, it may not be able to successfully complete the fracture reduction as expected. SUMMARY

[0006] In order to solve the problem that only using one reset path may not successfully complete the fracture reset as expected when reset error occurs, the fracture reset dynamic planning system based on three-dimensional reconstruction technology is provided, which not only obtains an optimal path, but also allocates multiple backup reset paths for the optimal path, so that when reset error occurs in the specific fracture reset process, a dynamic switching mechanism based on multiple backup reset paths can be established, and error tolerance can be realized through path switching index optimization.

[0007] The fracture reset dynamic planning system based on three-dimensional reconstruction technology adopts the following technical scheme:

[0008] One embodiment of the present application provides a fracture reset dynamic planning system based on three-dimensional reconstruction technology, which comprises the following modules:

[0009] The three-dimensional reconstruction module is used for three-dimensional reconstruction of a muscle-bone model, and the muscle-bone model at least comprises a bone model and a muscle model; a fractured bone in the muscle-bone model is recorded as a child bone and a parent bone; and other bones and muscles outside the child bone are recorded as constraint elements interfering with the fracture reset process.

[0010] The path acquisition module is used for acquiring candidate paths in which the child bone does not collide with the constraint elements when moving to the parent bone in the fracture reset process on the muscle-bone model.

[0011] Each candidate path comprises a bone posture of each position of the child bone in the moving process; and the constraint element posture of the child bone at each position on each candidate path is simulated.

[0012] The path optimization module is used for recording any one candidate path as a target path, and recording any one position of the target path as a target position.

[0013] On all candidate paths except the target path, a position most similar to the target position and having a most similar bone posture is acquired and recorded as a reference position of the target position, and a candidate path in which the reference position is located is recorded as a reference path of the target path.

[0014] The similarity between the constraint element posture of the position before the target position on the target path and the constraint element posture of the position before the reference position on the reference path is recorded as a path switching index of the target position; a target position with a path switching index greater than a preset threshold th is recorded as a path switching point, and a candidate path in which the reference position of the path switching point is located is recorded as a switching path of the target path.

[0015] A target path with the maximum sum of path switching indexes of all path switching points and the most path switching points on the switching path is recorded as an optimal path, and all switching paths of the optimal path are used as backup reset paths when reset error occurs in the fracture reset.

[0016] Preferably, the specific steps for obtaining the position that is closest to the target position and has the most similar skeletal pose among all candidate paths outside the target path, and denoting it as the reference position of the target position, are as follows:

[0017] For any candidate path among all candidate paths other than the target path, obtain the Euclidean distance 'a' between any position on the candidate path and the target position, and the similarity 'b' between the skeletal pose of any position and the skeletal pose of the target position; define the first index of any position on the candidate path as: ba;

[0018] On any candidate path, obtain the position with the largest first indicator and record it as the candidate position; among all candidate paths other than the target path, select the candidate position with the largest first indicator as the reference position of the target position.

[0019] Preferably, the similarity between the poses of the constraint elements at positions preceding the target position on the target path and the poses of the constraint elements at positions preceding the reference position on the reference path is denoted as the path switching index for the target position, and the specific steps include the following:

[0020] Each constraint element at each position on each candidate path corresponds to an attitude vector;

[0021] The similarity between the attitude vector sequence formed by the attitude vectors of each constraint element at all positions before the target position on the target path and the attitude vector sequence formed by the attitude vectors of each constraint element at all positions before the reference position on the reference path is denoted as the attitude change similarity of each constraint element; the mean of the attitude change similarities of all constraint elements is denoted as the path switching index of the target position.

[0022] Preferably, the target path that has the largest sum of path switching indices for all path switching points and the most path switching points on the switching path is denoted as the optimal path, and the specific steps included are as follows:

[0023] The sum of the path switching indices of all path switching points on each target path is recorded as the first feature of each target path. For each switching path of each target path, the number of path switching points on each switching path is recorded, and the average number of path switching points on all switching paths of each target path is recorded as the second feature. The sum of the first feature and the second feature of each target path is recorded as the optimal index of each target path, and the target path with the largest optimal index is taken as the optimal path.

[0024] Preferably, the specific steps for obtaining the similarity of the skeletal poses are as follows:

[0025] The rotation angles of the skeleton pose around each coordinate axis in the three-dimensional space where the skeleton model is located form a vector, and the cosine similarity of the vectors corresponding to any two skeleton poses is taken as the similarity of the skeleton poses.

[0026] Preferably, the specific acquisition steps of the pose vector are as follows:

[0027] When each sub-bone is taken as a constraint element, the position of the skeleton and the rotation angles of the skeleton around each coordinate axis in the three-dimensional space where the skeleton model is located are taken as the pose vector.

[0028] When each muscle is taken as a constraint element, the stretch length and volume of each muscle and the position of each muscle are taken as the pose vector.

[0029] Preferably, the specific acquisition steps of the pose change similarity are as follows:

[0030] PCA dimension reduction is performed on all the pose vectors in each pose vector sequence, each pose vector is reduced to one dimension, and a sequence formed by the dimension reduction results of all the pose vectors in each pose vector sequence is recorded as a dimension reduction sequence.

[0031] The DTW distance of the dimension reduction sequences of any two pose vector sequences is recorded as y, and exp(-y) is recorded as the pose change similarity, where exp() represents an exponential function with a natural constant as the base.

[0032] Preferably, the specific steps of taking all the switching paths of the optimal path as the backup reduction paths when the reduction error occurs in the fracture reduction include:

[0033] In the fracture reduction using the optimal path, the real-time detected position of the sub-bone and the real-time detected skeleton pose are recorded as a real-time position and a real-time pose respectively, and the position where the reduction error occurs is obtained according to the differences between the real-time position and the real-time pose and the positions and the skeleton poses on the optimal path, and is recorded as a reduction error position.

[0034] Among all the path switching points on the optimal path, the path switching point closest to the reduction error position is obtained, and when the distance between the reduction error position and the closest path switching point is less than a third preset threshold th3, the sub-bone is moved from the reduction error position to the reference position of the closest path switching point, and then the fracture reduction is performed according to the backup reduction path where the reference position is located.

[0035] Preferably, the specific steps of obtaining the position where the reduction error occurs according to the differences between the real-time position and the real-time pose and the positions and the skeleton poses on the optimal path, and recording the position as a reduction error position, include:

[0036] Obtaining a position closest to the real-time position on the optimal path, the distance between the real-time position and the position is recorded as ra, obtaining the similarity rb between the real-time posture and the posture of the position on the optimal path, and determining that there is a reset error when rb is less than a first preset threshold th1 and ra is greater than a second preset threshold th2, and the position is used as a reset error position.

[0037] Preferably, when the distance between the reset error position and the nearest path switching point is greater than or equal to th3, the value of the preset threshold th is reduced multiple times, and after each time the value of the preset threshold th is reduced, the target position on the optimal path with the path switching index greater than the preset threshold th is recorded as the path switching point again; until the distance between the nearest path switching point on the optimal path and the reset error position is not greater than or equal to th3, the value of the preset threshold th is stopped being reduced; and then the sub-bone is moved from the reset error position to the reference position of the nearest path switching point, and the fracture reset is performed according to the standby reset path where the reference position is located.

[0038] The beneficial effects of the technical scheme of the present application are:

[0039] The optimal path obtained by the present application can ensure that when a reset error occurs, on the one hand, it can reliably switch to an alternative fracture reset path, and the reliability refers to that the optimal path as a whole has a larger path switching index, which is obtained from the position and posture of the sub-bone and the posture change of the constraint element, ensuring that the sum of the path switching indexes of all path switching points is maximum, so that even if multiple reset errors occur, there is still an opportunity to switch to the standby reset path, or the risk of switching to the alternative reset path can be avoided as much as possible when switching; on the other hand, if there is still a reset error after switching to the standby reset path, it is still possible to switch to other standby reset paths as much as possible.

[0040] Compared with the case where only one reset path is used and the fracture reset cannot be successfully completed as expected when a reset error occurs, the optimal path obtained by the present application has multiple reliable standby reset paths, which can ensure that the fracture reset process is successfully completed, and error tolerance is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the drawings needed in the following embodiment or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0042] Figure 1A framework structure diagram of a fracture reduction dynamic planning system based on a three-dimensional reconstruction technique according to an embodiment of the present application is shown in FIG. 1.

[0043] Figure 2 A flowchart of a fracture reduction path switching method in a fracture reduction dynamic planning system based on a three-dimensional reconstruction technique according to an embodiment of the present application is shown in FIG. 6. DETAILED DESCRIPTION

[0044] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined purposes, the specific embodiments, structures, features and effects of the fracture reduction dynamic planning system based on a three-dimensional reconstruction technique according to the present application are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0046] The specific scheme of the fracture reduction dynamic planning system based on a three-dimensional reconstruction technique according to the present application is described in detail below in combination with the accompanying drawings.

[0047] Embodiment One:

[0048] Please refer to Figure 1 which shows a framework structure diagram of a fracture reduction dynamic planning system based on a three-dimensional reconstruction technique according to an embodiment of the present application. The system includes a three-dimensional reconstruction module, a path acquisition module and a path optimization module.

[0049] Specifically, the three-dimensional reconstruction module is used for three-dimensional reconstruction of a musculoskeletal model. Specifically, according to CTA images, three-dimensional reconstruction based on the Marching-Cubes algorithm is adopted for bone and skin models, and preoperative MRI (magnetic resonance examination) is used to detect nerve distribution; CTA-MRI image fusion and registration technology is adopted to realize three-dimensional geometric reconstruction of important tissues such as bones, muscles, blood vessels and nerves. This process is a known technology, for example, a lower limb bone high-energy injury reduction navigation method disclosed in a patent with publication number CN116370073A. The musculoskeletal model in this embodiment includes bone, muscle, blood vessel and nerve models.

[0050] Additionally, the fracture reduction process follows the principle of "seeking the mother from the child", that is, moving the distal fracture end (child bone) to the proximal fracture end (mother bone). The bones outside the child bone, as well as muscles, blood vessels and nerves, etc. are recorded as constraint elements that may block or interfere with the movement of the child bone to the mother bone; for example, the bones, blood vessels and muscles between the child bone and the mother bone will block or interfere with the movement of the child bone to the mother bone, for another example, the muscles attached to the child bone will also interfere with the process; when the child bone collides with these bones or blood vessels during the movement process, or stretches these muscle tissues, it will leave sequelae for the fracture reduction or interfere with the postoperative rehabilitation process.

[0051] A path acquisition module is configured to acquire candidate paths on the musculoskeletal model, which are not collided with the constraint elements when the child bone moves to the mother bone in the fracture reduction process.

[0052] As an example, the method for acquiring candidate paths includes:

[0053] Medical professionals in the field of orthopedics mark the movement path of the child bone to the mother bone on the musculoskeletal model, delete the movement path collided with the constraint elements or the movement path determined by the medical professionals as inappropriate, and the remaining movement path is recorded as a candidate path.

[0054] As another example, the method for acquiring candidate paths includes:

[0055] For the patent with publication number CN116370073A, a lower limb skeletal high-energy injury reduction navigation method is disclosed, which constructs a reduction path expression based on collision detection of bones, blood vessels and nerves, and solves a series of reduction basic paths based on this, and simultaneously simulates and calculates the muscle force (i.e. the resultant force of the muscle when deformed). In this embodiment, the reduction basic path in which the muscle force is not greater than the maximum bearing force of each muscle in the series of reduction basic paths is recorded as a candidate path.

[0056] In other examples, the method for acquiring candidate paths includes:

[0057] The artificial force field based method plans the moving path of the sub-bone, which uses a virtual force field to guide the object (i.e. the sub-bone) to avoid obstacles (i.e. the constraint elements). For example, a repulsive force field is created between the sub-bone and other bones or muscles; specifically, the product of the shortest distance between the sub-bone and other bones (or muscles) and a first preset parameter (e.g. 1) is taken as the repulsive force field. For another example, a muscle-bone force field is created between the sub-bone and the muscle connected thereto; specifically, the product of the stretching amount of each muscle and a second preset parameter (e.g. 0.5) is taken as a first muscle-bone force field, and the product of the twisting degree of the muscle and a third preset parameter (e.g. 0.5) is taken as a second muscle-bone force field. The resultant force field of the repulsive force field, the first muscle-bone force field and the second muscle-bone force field is regarded as the virtual force field. The artificial force field based method plans the path based on the virtual force field.

[0058] The method of planning the path based on the virtual force field by using the artificial force field based method is a known technology, and the present embodiment will not be described in detail. In the present embodiment, the first preset parameter, the second preset parameter and the third preset parameter are respectively set to different values, so that the artificial force field based method outputs a plurality of moving paths, which are recorded as candidate paths.

[0059] In some examples, the method of obtaining the candidate paths includes:

[0060] The candidate paths obtained by all the above examples are taken as a set, and a plurality of (e.g. 20) candidate paths are selected from the set by using the expert scoring method as the candidate paths obtained by the present example.

[0061] Thus, a plurality of candidate paths are obtained.

[0062] The candidate paths described in the present embodiment refer to all positions (e.g. one position is sampled every 1 mm) of the sub-bone during the movement of the sub-bone to the parent bone, and each position is three-dimensional in the muscle-bone model. Meanwhile, the sub-bone corresponds to a bone posture at each position, which represents the posture of the sub-bone at the position. The bone posture in the present embodiment is the rotation angle of the sub-bone in the three-dimensional space of the muscle-bone model (referred to as the rotation angle of the sub-bone).

[0063] In summary, the candidate paths describe the translation and rotation at each position during the movement of the sub-bone.

[0064] Further, the constraint element posture of the sub-bone at each position on each candidate path is simulated; specifically:

[0065] When the sub-bone moves on each candidate path, it directly or indirectly affects the posture of the constraint element.

[0066] Simulate the pose of the constraint element at each position (i.e., the constraint element pose at each position) when the sub-bone moves to each position (i.e., from the last position to the current position).

[0067] As an example, when each other bone (other than the sub-bone) is the constraint element, the position of each other bone (specifically, the center position of the bone model) and the rotation angle of each other bone are the constraint element pose at each position.

[0068] As an example, when each muscle is the constraint element, the muscle force of each muscle (i.e., the resultant force of the muscle when deformed) and the position of each muscle model are the constraint element pose at each position.

[0069] In other embodiments, the stretch length or volume of each muscle and the position of each muscle model are the constraint element pose at each position.

[0070] As an example, when a blood vessel or a nerve is the constraint element, the blood vessel or nerve needs to be divided into segments in advance in the musculoskeletal model, for example, one segment per centimeter, and then each segment of the blood vessel or nerve model is a constraint element. The length, position, and rotation angle of each segment of the blood vessel or nerve model are the constraint element pose at each position.

[0071] In other embodiments, the extrusion deformation variable (e.g., the deformation variable of the average diameter or volume) of each segment of the blood vessel or nerve model is the constraint element pose at each position.

[0072] As an example, the specific method of simulating the pose of the constraint element at each position (i.e., the constraint element pose at each position) includes:

[0073] The Hill three-element model (shrinkage unit, parallel elastic unit, and serial elastic unit) is used to describe the muscle dynamics, and in platforms such as Unity, the Rigidbody component and the Character Joint joint system are used to simulate the pose of the constraint element at each position.

[0074] In other embodiments, the Ziva VFX soft tissue dynamics engine can be used to simulate the pose of the constraint element at each position.

[0075] In some embodiments, the finite element analysis method can be used to simulate the pose of the constraint element at each position.

[0076] The above methods are known techniques, and the present embodiment does not go into specific details.

[0077] The path optimization module is configured to filter out a candidate path from all candidate paths as an optimal path of the fracture reduction (referred to as the optimal path for short).

[0078] The module considers that, in the process of fracture reduction, whether it is manual fracture reduction or surgical navigation fracture reduction by using a robot, due to the complexity of human tissues, there may be a reduction error when the fracture reduction path designed before the operation is followed, which means that the position and posture of the sub-bone do not change along the fracture reduction path during the sub-bone reduction process, but deviate from the designed fracture reduction path; this leads to the fact that the subsequent fracture reduction process needs to rely on the experience of doctors to complete, or the fracture reduction result deviates significantly from the expectation, or even when the fracture reduction is continued according to the fracture reduction path designed before the operation, it will obviously cause secondary damage to the patient.

[0079] The embodiment considers the above problems when designing or planning the optimal path, so that there is still a usable backup reduction path when there is a reduction error in the reduction process, and the above problems caused by always relying on a single fracture reduction path are avoided as much as possible.

[0080] Specifically:

[0081] (1) Any one candidate path is recorded as a target path, and any one position on the target path is recorded as a target position; on all candidate paths other than the target path, the position closest to the target position and having the most similar bone posture is obtained, which is recorded as the reference position of the target position, and the candidate path on which the reference position is located is recorded as the reference path of the target path.

[0082] Wherein, the reference position on the reference path and the target position on the target path are closest, and have the same and similar bone postures; then when the sub-bone is reduced along the target path, the position of the sub-bone and its bone posture are monitored in real time, if the sub-bone does not move towards the target position, or the real-time monitored bone posture of the sub-bone is significantly different from the bone posture of the target position, at this time it indicates that the sub-bone deviates from the target path at the target position, at this time the reference path can be used as a backup reduction path for fracture reduction, for reference in the subsequent fracture reduction process.

[0083] As an example, on all candidate paths other than the target path, the position closest to the target position and having the most similar bone posture is obtained, which is recorded as the reference position of the target position, including the method of:

[0084] For any one candidate path in all candidate paths other than the target path, the Euclidean distance a between any one position on the candidate path and the target position is obtained, and the similarity b between the bone posture of any one position and the bone posture of the target position is obtained.

[0085] The first index of any position on the candidate path is defined as: b-a; the greater the first index, the closer the position to the target position and the more similar the bone posture.

[0086] The position with the largest first index on any candidate path is obtained, which is recorded as the candidate position.

[0087] Similarly, for all candidate paths outside the target path, each candidate path corresponds to a candidate position, and the candidate position with the largest first index is again selected as the reference position of the target position.

[0088] In some embodiments, b x w1-a x w2 can be used as the first index. Wherein w1 and w2 are preset attention coefficients, some embodiments take w1 and w2 equal to 0.1 and 1 as examples for description, and other embodiments can set w1 and w2 to other values.

[0089] As an example, the calculation method of the similarity of the bone posture includes:

[0090] Since the bone posture contains multiple numerical values (such as rotation angles around each coordinate axis in three-dimensional space), these numerical values form a vector, and the cosine similarity of the vectors corresponding to any two bone postures is used as the similarity of the bone postures.

[0091] In other embodiments, the Euclidean distance x of the vectors corresponding to any two bone postures can be used, and exp(-x) can be used as the similarity of the bone postures; exp() represents the exponential function with the natural constant as the base.

[0092] (2) The similarity of the constrained element postures at the positions before the target position on the target path and the positions before the reference position on the reference path, recorded as the path switching index of the target position; the target position with a path switching index greater than a preset threshold is recorded as the path switching point, and the candidate path where the reference position of the path switching point is located is recorded as the switching path of the target path.

[0093] The embodiment further considers that the reference position on the reference path and the target position on the target path only describe the difference or similarity of the sub-bone positions and the bone posture, and the reference path obtained at this time cannot necessarily be used as a better backup reset path when the target path has a reset error. The path switching index in the embodiment further describes the similarity between the constraint element postures when the sub-bone moves according to the target path and the reference path, and the path switching point and the switching path obtained based on the path switching index can be more reliable as a better backup reset path when the target path has a reset error; wherein the greater the path switching index, the more similar the reference path is to the target path in terms of the similar sub-bone postures at the reference position and the target position, and the similar influence and interference on the constraint element, at this time, the reference path can be used as a backup reset path when the target path has a reset error, or even as a replacement for the target path.

[0094] As an example, the similarity between the constraint element postures at the positions before the target position on the target path and the constraint element postures at the positions before the reference position on the reference path is recorded as the path switching index of the target position, which includes:

[0095] First of all, it needs to be pointed out that for any position on any candidate path, the constraint element posture at this position describes the posture of all constraint elements simulated when the sub-bone moves along the candidate path. Each constraint element posture also contains multiple values (see the path acquisition module for details), which form the posture vector of the constraint element. That is, each constraint element at each position on each candidate path corresponds to a posture vector.

[0096] For all positions before the target position on the target path (including the target position), the posture vectors of each constraint element at all positions form a posture vector sequence, which reflects the posture change of each constraint element before the target position on the target path.

[0097] For all positions before the reference position on the reference path (including the reference position), the posture vectors of each constraint element at all positions also form a posture vector sequence, which reflects the posture change of each constraint element before the reference position on the reference path.

[0098] For the same constraint element, the similarity between the posture vector sequence obtained for the constraint element on the target path and the posture vector sequence obtained for the constraint element on the reference path is recorded as the posture change similarity of each constraint element. The average of the posture change similarities of all constraint elements is recorded as the path switching index of the target position.

[0099] As an example, the method for obtaining the similarity of two pose vector sequences comprises:

[0100] First of all, it should be noted that the lengths of the two pose vector sequences are not necessarily the same.

[0101] PCA dimension reduction is performed on all the pose vectors in each pose vector sequence, each pose vector is reduced to one dimension, and the sequence formed by the dimension reduction results of all the pose vectors in each pose vector sequence is recorded as a dimension reduction sequence.

[0102] The DTW distance of the dimension reduction sequences of the two pose vector sequences is recorded as y, and exp(-y) is recorded as the similarity of the two pose vector sequences.

[0103] The DTW distance is obtained by the DTW algorithm, which is a known technology, and the present embodiment does not make specific limitations.

[0104] Further, the target position with a path switching indicator greater than a preset threshold th is recorded as a path switching point, and the candidate path in which the reference position of the path switching point is located is recorded as a switching path of the target path.

[0105] The present embodiment takes th=0.3 as an example for description, and other embodiments can set it to other values, and the present embodiment does not make specific limitations.

[0106] At this point, several path switching points can be obtained on any one candidate path.

[0107] (3) The path switching points on the target path are obtained as described above, and a path switching indicator is obtained for each path switching point. For all target paths, the sum of the path switching indicators of all path switching points on each target path is obtained, which is recorded as a first feature of each target path. The greater the first feature, the more reliable the switching from the target path to the alternative fracture reduction path when there is a reset error.

[0108] On the other hand, for each switching path of each target path, each switching path also has several path switching points, and the number of path switching points is recorded. The average number of path switching points on all switching paths of each target path is recorded as a second feature, which represents the average path switching point on the switching path of each target path. The greater the second feature, the more path switching points the switching path of the target path has.

[0109] The sum of the first feature and the second feature of each target path is recorded as a preferred indicator of each target path, and the target path with the largest preferred indicator is taken as the path of fracture reduction (hereinafter referred to as the optimal path). All switching paths of the optimal path are used as backup reduction paths.

[0110] The optimal path obtained above can ensure that when a reset error occurs, on the one hand, the alternative fracture reset path can be reliably switched to, and the reliability refers to that the optimal path as a whole has a larger path switching index obtained from the position of the sub-bone, the bone posture and the posture change of the constraint element, so that the first feature and the second feature have maximum values at the same time, so that there is an opportunity to switch to the backup reset path even when multiple reset errors occur, or the risk of switching to the alternative reset path can be avoided as much as possible when switching, for example, the risk of the fracture reset failing to be successfully completed and being not conducive to subsequent rehabilitation when switching due to the need for the sub-bone position to move a significant displacement or the need for the bone posture of the sub-bone to be significantly changed, or the posture of the interference element in the backup reset path cannot be adapted. On the other hand, when the reset error still exists after switching to the backup reset path, other backup reset paths are still switched to as much as possible.

[0111] Compared with the case of using only one reset path and possibly failing to successfully complete the fracture reset as expected when a reset error occurs, the optimal path obtained in the embodiment has multiple reliable backup reset paths, which can ensure that the fracture reset process is successfully completed.

[0112] Embodiment two:

[0113] The above embodiment one obtains the reset path of the fracture reset (i.e., the optimal path) and several alternative paths of the optimal path through the three-dimensional reconstruction technology before the fracture reset, and this process is completed before the fracture reset surgery.

[0114] Next, the fracture reset is performed according to the optimal path, and this process is well known; for example, manual reset according to the optimal path; for example, using a robot of a surgical navigation system to fix the sub-bone and automatically reset according to the optimal path; for example, a combination of manual reset and automatic reset.

[0115] The embodiment provides a method for obtaining a reset error in fracture reset, comprising:

[0116] The surface of the sub-bone is marked by an infrared optical tracker or an electromagnetic sensor, and the position and the bone posture of the sub-bone are obtained in real time, which are respectively denoted as a real-time position and a real-time posture. For example, a binocular vision method is used to track the position and the posture (i.e., the rotation angle) of a plurality of reflective ball markers fixed on the bone. The position closest to the real-time position on the optimal path is obtained, the distance between the position and the real-time position is denoted as ra, and the similarity rb between the bone posture of the position on the optimal path and the real-time posture is obtained. When rb is less than a first preset threshold th1 and ra is greater than a second preset threshold th2, it is indicated that the real-time position and the real-time posture deviate from the optimal path significantly, and at this time, it is determined that a reset error occurs at the position on the optimal path.

[0117] The embodiment takes th1=0.2 and th2=3 as an example, and th1 and th2 can be set to other values in other embodiments, which are not limited in the embodiment.

[0118] It should be noted that the unit of the Euclidean distance between the positions of the bones in all embodiments of the application is millimeter.

[0119] The embodiment provides a method, as shown in Figure 2 Switch to the standby reset path when a reset error occurs.

[0120] First of all, it should be noted that from the first embodiment: there are several path switching points on the optimal path, and each path switching point corresponds to a reference position, which is on the standby reset path.

[0121] When the fracture reset is performed by using the optimal path obtained above, the position where the reset error occurs is recorded as the reset error position. Among all the path switching points on the optimal path, the path switching point closest to the reset error position is obtained, and when the Euclidean distance between the closest path switching point and the reset error position is less than a third preset threshold th3, the sub-bone is directly moved from the reset error position to the reference position of the closest path switching point, and the trajectory of the movement is obtained by linear interpolation from the reset error position (and the bone posture at the position) and the reference position of the closest path switching point (and the bone at the position). After moving to the reference position of the closest path switching point, the subsequent fracture reset is performed according to the standby reset path where the reference position is located (at this time, the standby reset path is regarded as the optimal path).

[0122] The embodiment takes th3=0.5*th2 as an example, and 0.5 can be replaced with other values in other embodiments, preferably a value less than 1 and greater than or equal to 0, which is not limited in the embodiment.

[0123] In particular, when the Euclidean distance between the nearest path switching point and the reset error position is greater than or equal to the preset threshold th3, it indicates that there is no path switching point near the reset error position. At this time, the preset threshold th in the first embodiment is reduced (for example, it is reduced by 10% of the initial value, and the initial value of th is the value set in the first embodiment), and then the path switching points on the optimal path are obtained again (i.e., the target positions on the optimal path with the path switching index greater than the preset threshold th are recorded as the path switching points again), and then it is judged again whether the Euclidean distance between the nearest path switching point and the reset error position is greater than or equal to the preset threshold th3. If not, the standby reset path is switched according to the above method; if yes, the preset threshold th in the first embodiment is reduced again (for example, it is reduced by 10% of the initial value again), and the process is repeated until the Euclidean distance between the nearest path switching point and the reset error position is not greater than or equal to the preset threshold th3.

[0124] In particular, when the reduced preset threshold th is less than 0.1, the preset threshold th is no longer reduced, and the standby reset path is no longer switched, but the sub-bone is directly moved from the reset error position to the nearest position on the optimal path, and then the fracture reset is continued according to the optimal path obtained in the first embodiment.

[0125] When the fracture reset is performed only by using one optimal path (i.e., when the single-path fracture reset in the traditional method is used), in 100 simulation experiments, the success rate is 70%, i.e., in 30% of the simulation experiments, the fracture reset needs to rely on the experience of doctors, and in these 30% of the simulation experiments, the expected fracture reset effect is not achieved. In the 100 simulation experiments of the present embodiment, the success rate is 92%, and in 92% of the simulation experiments, the expected fracture reset effect is achieved. It can be seen that the dynamic switching of the optimal path and the standby reset path greatly improves the success rate of fracture reset. In the 100 simulation experiments of the present embodiment, the values of th and the number of path switching points are shown in Table 1. The average length of the optimal path in the 100 simulation experiments is 18 mm.

[0126] Table 1: Value of th and number of path switching points

[0127] value of th 0.3 0.27 0.24 0.21 0.18 0.15 0.12 number of path switch points 7.1 8.7 9.4 12.7 13.2 13.6 15.5 number of simulation experiments 28 17 15 10 13 11 6

[0128] For example, the second table in Table 1 indicates that in 28 simulation experiments, the final value of th is set to 0.3, at this time, 7.1 path switching points can be obtained on average in each path of the optimal path and all standby reset paths. According to Table 1, in most simulation experiments, the fracture reset is completed when th = 0.3.

[0129] The above merely provides the preferred embodiment of the present application, and is not used to limit the present application, and any modification, equivalent replacement, improvement, etc. made within the principle of the present application should be included in the protection scope of the present application.

Claims

1. A dynamic planning system for fracture reduction based on three-dimensional reconstruction technology, characterized in that, The system comprises the following modules: a three-dimensional reconstruction module for three-dimensionally reconstructing a musculoskeletal model, the musculoskeletal model comprising at least: a bone model and a muscle model; a fractured bone in the musculoskeletal model being recorded as a child bone and a parent bone; other bones and muscles outside the child bone being recorded as constraint elements interfering with the fracture reduction process; a path acquisition module for acquiring, on the musculoskeletal model, candidate paths in which the child bone moves to the parent bone without colliding with the constraint elements in the fracture reduction process; each candidate path comprising a bone posture of each position of the child bone in the moving process; and simulating constraint element postures of the child bone at each position on each candidate path; a path optimization module for recording any one candidate path as a target path, and any one position of the target path as a target position; acquiring, on all candidate paths other than the target path, a position closest to the target position and having a most similar bone posture as the reference position of the target position, and recording the candidate path on which the reference position is located as the reference path of the target path; similarity between constraint element postures of positions before the target position on the target path and constraint element postures of positions before the reference position on the reference path being recorded as a path switching indicator of the target position; recording a target position with the path switching indicator greater than a preset threshold th as a path switching point, and recording the candidate path on which the reference position of the path switching point is located as a switching path of the target path; acquiring a target path with a maximum sum of path switching indicators of all path switching points and a maximum number of path switching points on the switching path, and recording the target path as an optimal path, and recording all switching paths of the optimal path as backup reduction paths in the event of a reduction error in the fracture reduction.

2. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 1, wherein, The specific steps of acquiring, on all candidate paths other than the target path, a position closest to the target position and having a most similar bone posture as the reference position of the target position, include the following: for any one candidate path among all candidate paths other than the target path, acquiring a Euclidean distance a between any one position on the candidate path and the target position, and a similarity b between a bone posture of any one position and a bone posture of the target position; and defining a first indicator of any one position on the candidate path as: b-a; acquiring a position with a maximum first indicator on the any one candidate path, and recording the position as a candidate position; and selecting the candidate position with the maximum first indicator among all candidate paths other than the target path as the reference position of the target position.

3. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 1, wherein, The specific steps of recording similarity between constraint element postures of positions before the target position on the target path and constraint element postures of positions before the reference position on the reference path as a path switching indicator of the target position, include the following: each constraint element at each position on each candidate path corresponds to a posture vector; a posture vector sequence formed by posture vectors of each constraint element at all positions before the target position on the target path is similar to a posture vector sequence formed by posture vectors of each constraint element at all positions before the reference position on the reference path, and the similarity is recorded as a posture change similarity of each constraint element; and the mean value of the posture change similarities of all constraint elements is recorded as the path switching indicator of the target position.

4. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 1, wherein, The target path with the maximum sum of path switching indicators of all path switching points and the maximum number of path switching points on the switching path is recorded as an optimal path, and the specific steps include the following: For all target paths, the sum of path switching indicators of all path switching points on each target path is obtained, and recorded as the first feature of each target path; For each switching path of each target path, the number of path switching points on each switching path is recorded, and the average number of path switching points on all switching paths of each target path is recorded as the second feature; The sum of the first feature and the second feature of each target path is recorded as the preferred indicator of each target path, and the target path with the maximum preferred indicator is taken as the optimal path.

5. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 2, wherein, The specific steps for obtaining the similarity between the bone posture of the arbitrary position and the bone posture of the target position are as follows: The rotation angles included in the bone posture form a vector, the rotation angles are the rotation angles of the bone around each coordinate axis in the three-dimensional space where the musculoskeletal model is located, and the cosine similarity of the vectors corresponding to any two bone postures is taken as the similarity of the bone postures.

6. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 3, wherein, The specific steps for obtaining the posture vector are as follows: When each sub-bone is taken as a constraint element, the position of the bone and the rotation angles of the bone around each coordinate axis in the three-dimensional space where the musculoskeletal model is located are taken as the posture vector; When each muscle is taken as a constraint element, the stretch length and volume of each muscle and the position of each muscle are taken as the posture vector.

7. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 3, wherein, The specific steps for obtaining the posture change similarity are as follows: PCA dimension reduction is performed on all posture vectors in each posture vector sequence, each posture vector is reduced to one dimension, and the sequence formed by the dimension reduction results of all posture vectors in each posture vector sequence is recorded as a dimension reduction sequence; The DTW distance of the dimension reduction sequences of any two posture vector sequences is recorded as y, and exp(-y) is recorded as the posture change similarity, where exp() represents the exponential function with the natural constant as the base.

8. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 1, wherein, The specific steps for taking all switching paths of the optimal path as the standby reset path when reset error occurs in the bone fracture reset include the following: During the bone fracture reset using the optimal path, the real-time position and real-time posture of the sub-bone are recorded as the real-time position and real-time posture respectively; the position where the reset error occurs is obtained according to the difference between the real-time position and real-time posture and the position and bone posture on the optimal path, and recorded as the reset error position; Among all path switching points on the optimal path, the path switching point closest to the reset error position is obtained; when the distance between the reset error position and the closest path switching point is less than a third preset threshold th3, the sub-bone is moved from the reset error position to the reference position of the closest path switching point, and then the bone fracture reset is performed according to the standby reset path where the reference position is located.

9. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 8, wherein, The specific steps for obtaining the position where the reset error occurs according to the difference between the real-time position and real-time posture and the position and bone posture on the optimal path, and recording the position as the reset error position, include the following: Obtaining a position closest to the real-time position on the optimal path, the distance between the real-time position and the position being recorded as ra, obtaining a similarity rb between the real-time pose and the skeleton pose of the position on the optimal path, and determining that the reset error exists when the rb is less than a first preset threshold th1 and the ra is greater than a second preset threshold th2, the position being used as the reset error position.

10. The system for dynamic planning of fracture reduction based on three-dimensional reconstruction technology according to claim 8, wherein, When the distance between the reset error position and the nearest path switching point is greater than or equal to th3, the value of the preset threshold th is reduced multiple times, and after each reduction of the value of the preset threshold th, the target position on the optimal path for which the path switching index is greater than the preset threshold th is recorded as a path switching point again; until the distance between the nearest path switching point on the optimal path to the reset error position and the reset error position is not greater than or equal to th3, the reduction of the value of the preset threshold th is stopped; then, after moving the sub-bone from the reset error position to the reference position of the nearest path switching point, the fracture reduction is performed according to the standby reset path where the reference position is located; wherein .

Images