A method and apparatus for predicting a surgical indicator

Abstract

The specification discloses a method and device for predicting a surgical index. The method comprises the following steps: acquiring a full-length image of a target limb of a patient and a three-dimensional bone model of a surgical region of the target limb; reconstructing a three-dimensional bone model in which a bone in the three-dimensional bone model is replaced by a joint prosthesis; adjusting the reconstructed three-dimensional bone model according to an expected bone relationship; and predicting, based on a projection result of the adjusted three-dimensional bone model in the full-length image and the full-length image, a surgical index for evaluating a postoperative recovery state of the target limb after the joint prosthesis is replaced. The method uses the three-dimensional bone model of the surgical region as a medium, projects the prosthesis to the full-length image, and predicts the surgical index after the joint prosthesis is replaced based on the full-length image and the projection result, without the need for multiple-angle shooting of the target limb based on an X-ray generating device, so that the patient is subjected to a smaller dose of radiation and less damage.

Description

Technical Field

[0001] This manual relates to the field of medical technology, and in particular to a method and apparatus for predicting surgical indicators. Background Technology

[0002] Artificial joint replacement surgery involves using materials such as metal, high-molecular-weight polyethylene, and ceramics to create artificial joint prostheses tailored to the shape, structure, and function of human joints. These prostheses are surgically implanted into the body to replace the function of the diseased joint, thereby relieving joint pain and restoring joint function. Predictive surgical indicators are widely used in artificial joint replacement surgery because they can predict the stress on the patient's limb after the joint prosthesis is implanted.

[0003] A commonly used method for predicting surgical indicators is based on a full-length three-dimensional skeletal model of the target limb obtained preoperatively. Specifically, a three-dimensional skeletal model of the target limb involved in the replacement joint prosthesis (e.g., a three-dimensional skeletal model of the lower limb region) is obtained, and the acquired three-dimensional skeletal model is segmented to determine the bone type and structure of each bone contained in the model. Then, based on the pre-designed three-dimensional models of each joint prosthesis and the determined bone type and structure, each joint prosthesis is simulated for implantation. Finally, medical professionals observe the implantation results to determine the surgical indicators corresponding to each joint prosthesis.

[0004] However, existing technologies require the determination of a three-dimensional skeletal model of the target limb involved in the replacement of the joint prosthesis. The three-dimensional skeletal model of the target limb is large in scope and usually requires X-ray generators to take pictures of the target limb from multiple angles. Based on the pictures, a full-length three-dimensional skeletal model of the target limb is reconstructed. This results in a large radiation dose for the patient, causing significant harm. Summary of the Invention

[0005] This specification provides a method and apparatus for predicting surgical indicators, thereby partially solving the aforementioned problems existing in the prior art.

[0006] The following technical solution is adopted in this specification:

[0007] This manual provides a method for predicting surgical parameters, including:

[0008] Acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb;

[0009] After replacing the bones in the three-dimensional skeletal model with joint prostheses, the three-dimensional skeletal model of the surgical area is reconstructed, and the reconstructed three-dimensional skeletal model is adjusted according to the expected bone positional relationship.

[0010] Based on the projection of the adjusted three-dimensional skeletal model onto the full-length image and the full-length image, surgical indicators are predicted after the replacement of the joint prosthesis. These surgical indicators are used to assess the postoperative recovery status of the target limb.

[0011] Optionally, based on the projection result of the adjusted three-dimensional skeletal model into the full-length image and the full-length image, surgical parameters after replacing the joint prosthesis are predicted, specifically including:

[0012] The adjusted 3D skeletal model is projected onto the full-length image, and the projection result is determined;

[0013] Based on the projection results, the positions of the bones in the full-length image are adjusted;

[0014] Based on the adjusted full-length image, surgical parameters after joint prosthesis replacement are predicted.

[0015] Optionally, the reconstructed 3D skeletal model may be adjusted according to the expected skeletal positional relationships, specifically including:

[0016] Determine the expected skeletal positional relationship corresponding to the joint prosthesis;

[0017] Based on the expected bone position relationship, the bone positions of at least some of the bones corresponding to the joint prosthesis in the reconstructed three-dimensional bone model are adjusted so that the position relationship of each bone in the adjusted three-dimensional bone model is the expected bone position relationship.

[0018] Optionally, determining the projection result of the adjusted 3D skeletal model into the full-length image specifically includes:

[0019] The full-length image and the adjusted 3D skeletal model are registered to determine the correspondence between the bones in the adjusted 3D skeletal model and the full-length image.

[0020] Based on the correspondence, the adjusted 3D skeletal model is projected onto the full-length image to determine the projection result.

[0021] Optionally, determining the correspondence between the adjusted 3D skeletal model and the bones in the full-length image specifically includes:

[0022] Based on the full-length image and the three-dimensional skeletal model of the surgical area, the region of the surgical area in the full-length image is determined as the first region;

[0023] Based on the bones contained in the first region of the full-length image and the bones contained in the three-dimensional bone model, the correspondence between the adjusted three-dimensional bone model and the bones in the full-length image is determined.

[0024] Optionally, based on the full-length image and the three-dimensional skeletal model of the surgical area, the region of the surgical area within the full-length image is determined, specifically including:

[0025] Based on the acquisition posture of the full-length image and the three-dimensional skeleton model, determine the initial transformation relationship between the first reference frame where the three-dimensional skeleton model is located and the second reference frame where the full-length image is located;

[0026] Based on the initial transformation relationship, determine the initial projection result of the three-dimensional skeleton model in the second reference frame;

[0027] Based on the difference between the initial projection result and the full-length image, the transformation relationship is updated, and based on the updated transformation relationship, the region where the final projection result of the three-dimensional skeleton model is located in the second reference frame is determined as the region of the surgical area in the full-length image.

[0028] Optionally, based on the adjusted full-length image, surgical parameters after joint prosthesis replacement are predicted, specifically including:

[0029] Determine the bone type and bone position of each bone in the adjusted full-length image;

[0030] Based on the bone type and location of each bone, the lower limb force line and / or the center position of each bone in the adjusted full-length image are determined as surgical indicators.

[0031] Optionally, the joint prosthesis has multiple sets, and the method further includes:

[0032] For each group of joint prostheses, the matching degree of the group of joint prostheses is determined based on the surgical indicators for replacing the group of joint prostheses and the preset standard indicators;

[0033] Based on the matching degree of each group of joint prostheses, a joint prosthesis that matches the patient is determined.

[0034] Optionally, based on the surgical parameters and preset standard parameters for replacing this group of joint prostheses, the matching degree of this group of joint prostheses is determined, specifically including:

[0035] Based on the three-dimensional structure of this group of joint prostheses, the corresponding osteotomy amount for this group of joint prostheses is determined;

[0036] The similarity of the joint prostheses in this group is determined based on the difference between the surgical indicators and the preset standard indicators.

[0037] The matching degree of the joint prosthesis group is determined based on the osteotomy amount and the preset weight of the osteotomy amount, as well as the similarity and the preset weight of the similarity.

[0038] This specification provides a device for predicting surgical parameters, the device comprising:

[0039] The acquisition module is used to acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb;

[0040] The reconstruction module is used to reconstruct the three-dimensional bone model of the surgical area after replacing the bones in the three-dimensional bone model with joint prostheses, and to adjust the reconstructed three-dimensional bone model according to the expected bone position relationship.

[0041] The prediction module is used to predict surgical indicators after replacing the joint prosthesis based on the projection result of the adjusted three-dimensional skeletal model in the full-length image and the full-length image. The surgical indicators are used to assess the postoperative recovery status of the target limb.

[0042] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for predicting surgical indicators.

[0043] This specification provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for predicting surgical indicators.

[0044] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects:

[0045] In the method for predicting surgical indicators provided in this specification, a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb are acquired. The three-dimensional skeletal model after replacing the bones in the three-dimensional skeletal model with a joint prosthesis is reconstructed. The reconstructed three-dimensional skeletal model is adjusted according to the expected skeletal relationship. Based on the projection result of the adjusted three-dimensional skeletal model in the full-length image and the full-length image, surgical indicators for assessing the postoperative recovery status of the target limb after replacing the joint prosthesis are predicted.

[0046] As can be seen from the above method, this method uses a three-dimensional skeletal model of the surgical area as a medium to project the prosthesis onto a full-length image, and predicts surgical indicators after joint replacement based on the full-length image and projection results. It does not require multi-angle imaging of the target limb based on an X-ray generator, resulting in a lower radiation dose to the patient and less harm to the patient. Attached Figure Description

[0047] The accompanying drawings, which are included to provide a further understanding of this specification and form part of this specification, illustrate exemplary embodiments and are used to explain this specification, but do not constitute an undue limitation thereof. In the drawings:

[0048] Figure 1 This is a flowchart illustrating the method for predicting surgical indicators provided in this manual.

[0049] Figure 2 This is a schematic diagram of the reconstructed three-dimensional skeletal structure provided in this specification;

[0050] Figure 3 A schematic diagram illustrating adjustments made to the 3D skeletal model provided in this manual;

[0051] Figure 4 This is a schematic diagram illustrating the determination of adjustments to the full-length image provided in this specification;

[0052] Figure 5 A schematic diagram illustrating the determination of surgical parameters provided in this instruction manual;

[0053] Figure 6 The device for predicting surgical parameters provided in this instruction manual;

[0054] Figure 7 The corresponding information provided in this specification Figure 1 A schematic diagram of an electronic device. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0056] Unlike existing technologies that acquire a full-length 3D skeletal model of the patient's target limb and then simulate implantation based on the bone types and structures within that model, as well as the 3D structure of the prosthesis, to determine surgical parameters for the joint prosthesis, this method involves a higher radiation dose to the patient, resulting in greater harm. This specification provides a novel method for predicting surgical parameters. Based on a full-length image of the patient's target limb, a 3D skeletal model of the surgical area, and pre-defined joint prostheses, this method predicts the surgical parameters for replacing the joint prosthesis. Compared to acquiring the 3D skeletal structure of the target limb, this method only requires acquiring a 3D skeletal model of the surgical area, resulting in a lower radiation dose and less harm to the patient.

[0057] The aforementioned surgical indicators are used to assess the postoperative recovery status after replacing the aforementioned joint prosthesis, so as to accurately predict the postoperative recovery after performing artificial joint replacement surgery, thereby improving the success rate of artificial joint replacement surgery and patient satisfaction. This solves the problem that in the existing technology, when the surgical indicators need to be manually observed to determine the implantation results, the determined surgical indicators may have errors.

[0058] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0059] Figure 1 This is a flowchart illustrating the method for predicting surgical indicators provided in this manual, which specifically includes the following steps:

[0060] S100: Acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb.

[0061] The method for predicting surgical indicators described in this manual can be applied before or during surgery, before replacing the prosthesis. It matches pre-stored joint prostheses with the patient's surgical site, thereby selecting and replacing the prosthesis. The specific execution time of this method for predicting surgical indicators can be set as needed, and this manual does not impose any restrictions on it.

[0062] The method for predicting surgical parameters described in this specification is applied to scenarios involving total joint replacement surgery. Since total joint replacement surgery involves implanting an artificial joint prosthesis into the patient's body to replace the function of the diseased joint, it is necessary to predict the surgical parameters of the patient's target limb after the replacement of the joint prosthesis, based on the skeletal structure of the target limb and a pre-defined 3D model of the prosthesis. In this method, where the full-length image of the target limb is a 2D image but the prosthesis is a 3D model, it is necessary to match the skeletal structure of the target limb with the prosthesis based on the 3D skeletal model of the surgical area. Therefore, the full-length image of the target limb and the 3D skeletal model of the target limb can be obtained first.

[0063] Specifically, a full-length image of the patient's target limb can be obtained using an X-ray generator. The target limb is the limb involved in the joint replacement surgery. For example, if the patient's surgical joint is the left knee, the target limb is the left lower limb, or a limb including the left lower limb. The full-length image of the target limb is obtained by stitching together multiple local images of the patient's target limb from the same direction obtained using the X-ray generator, resulting in a two-dimensional full-view image of the target limb. For example, an X-ray of the lower limb bones.

[0064] Then, a three-dimensional skeletal model of the target limb surgical area can be determined using computed tomography (CT). This surgical area is the region where surgery will be performed, determined before the actual surgery based on the patient's joint region; specifically, the joint area where the joint prosthesis will be replaced. The three-dimensional skeletal model of this target limb surgical area is a CT image of the joint that needs to be replaced. Of course, to ensure the accuracy of the acquired three-dimensional skeletal model during matching, a three-dimensional skeletal model of the area surrounding the patient's joint can also be obtained when acquiring the three-dimensional skeletal model of the target limb surgical area.

[0065] Furthermore, since this specification only requires acquiring the three-dimensional skeletal model of the surgical area of ​​the patient's target limb when obtaining the three-dimensional skeletal model, it is also possible to use a cone-beam 3D image acquisition device mounted on a C-arm to determine images of the patient's joint area from multiple angles. Based on the acquired images, a three-dimensional skeletal model of the patient's joint area can then be constructed to determine the three-dimensional skeletal model of the surgical area of ​​the patient's target limb.

[0066] Using the cone-beam 3D image acquisition device set on the C-arm as described in this manual to acquire a 3D skeletal model allows for the determination of the 3D skeletal model corresponding to the surgical site before joint replacement during the patient's surgery, eliminating the need for large-area irradiation of the patient using an X-ray generator.

[0067] Furthermore, the joint matching method described in this specification can be executed by electronic devices with computing capabilities, such as servers.

[0068] S102: After replacing the bones in the three-dimensional skeletal model with joint prostheses, reconstruct the three-dimensional skeletal model of the surgical area, and adjust the reconstructed three-dimensional skeletal model according to the expected bone position relationship.

[0069] In one or more embodiments provided in this specification, as previously described, it is necessary to determine the three-dimensional skeletal model of the surgical area after replacing the bones in the three-dimensional skeletal model with the joint prostheses based on the three-dimensional skeletal structure of the surgical area of ​​the patient's target limb and the three-dimensional structure of each joint prosthesis, so that the prosthesis and the two-dimensional full-length image can be matched.

[0070] Specifically, the server can determine the bone type and bone location of the bones contained in the three-dimensional bone model of the target limb surgical area determined in step S100.

[0071] Then, the server can replace the bones in the three-dimensional skeletal model of the surgical area with the set of joint prostheses based on the preset three-dimensional model of the joint prosthesis.

[0072] Finally, based on the replacement results, the server can reconstruct the 3D skeletal model of the surgical area, that is, determine the 3D skeletal model in which the corresponding bones are replaced with prostheses. For example... Figure 2 As shown.

[0073] Figure 2 This is a schematic diagram of the reconstructed three-dimensional skeletal structure provided in this specification. The left side of the diagram is a simplified schematic diagram of the three-dimensional skeletal model of the target limb surgical area. The server can determine the bone types and locations contained in this three-dimensional skeletal model, and replace the three-dimensional model of the pre-set joint prosthesis with the three-dimensional model of the prosthesis in the three-dimensional skeletal model, as shown on the right side of the diagram. The gray joint-shaped objects represent the joint prosthesis, and the white bone-shaped objects represent the bones contained in the three-dimensional skeletal model.

[0074] It should be noted that the type and number of bones included in this joint prosthesis can be set as needed, and this instruction manual does not impose any restrictions on this.

[0075] In addition, total knee replacement surgery involves replacing the patient's knee joint, which has deformities or other problems, with a joint prosthesis to relieve joint pain and restore knee joint mobility to the greatest extent possible. Maximizing knee joint mobility requires considering the stress distribution on the target limb. Therefore, the positions of the bones in the reconstructed 3D skeletal model need to be adjusted to the expected skeletal relationships to accurately determine surgical indicators such as the lower limb force line, which characterize the stress distribution on the target limb.

[0076] Based on this, the server can adjust the reconstructed 3D skeletal model according to the expected skeletal relationships.

[0077] Specifically, the server can pre-store the expected relationships between the bones involved in each joint. For example, in a knee joint prosthesis, the distance between the tibiofemoral joint and the patellar joint.

[0078] The server can then determine the relationship between the bones corresponding to a joint based on the joint identifier. The joint identifier can be manually entered or determined by the server based on the patient's user identifier.

[0079] Finally, the server can adjust the reconstructed 3D skeletal model based on the determined expected relationships, so that the relationships between the bones in the reconstructed 3D skeletal model are the expected skeletal relationships, etc. For example... Figure 3 As shown.

[0080] Figure 3 This diagram illustrates adjustments made to the 3D skeletal model provided in this manual. In the diagram, the white bone-like objects represent the bones in a simplified schematic of the 3D skeletal model, and the gray joint-like objects represent joint prostheses. The left side of the diagram shows the reconstructed 3D skeletal model, the middle side shows the expected bone positional relationships corresponding to the joint prosthesis, and the right side shows the 3D skeletal model adjusted based on the expected bone positional relationships.

[0081] Furthermore, due to the differences in shape and size among different prostheses, the expected skeletal relationships of different groups of prostheses are not entirely the same. Taking knee prostheses of the same shape but different sizes as an example, in larger knee prostheses, the gap between the femoral and tibial prostheses is larger than the gap between the smaller femoral and tibial prostheses. Therefore, different expected skeletal relationships can be preset for different prostheses.

[0082] Therefore, the server can first determine the expected skeletal positional relationship corresponding to this group of joint prostheses.

[0083] Then, the server can adjust the bone positions of at least some of the bones corresponding to the joint prosthesis in the reconstructed three-dimensional bone model according to the determined expected bone position relationship, so that the position relationship of each bone in the adjusted three-dimensional bone model is the expected bone position relationship.

[0084] Furthermore, by applying machine learning methods, when a 3D skeletal model is obtained, it can be used as input to output the corresponding bone positions and bone types, which is simple and fast. Therefore, machine learning methods can be used to determine the bone types and bone positions of the bones contained in the 3D skeletal model.

[0085] Specifically, the server can use the acquired 3D skeletal model as input to a pre-trained segmentation model to obtain the bone type and position of each bone in the 3D skeletal model output by the segmentation model.

[0086] The segmentation model can be trained in the following way:

[0087] First, three-dimensional skeletal models of several joint regions are obtained as first training samples, and the annotations of each first training sample are determined. The annotations of the first training samples include the bone type and bone location of each bone.

[0088] Then, each first training sample is used as input to the segmentation model to be trained, and the segmentation result of the first training sample output by the segmentation model is obtained.

[0089] Finally, based on the segmentation results and annotations of each first training sample, the first loss is determined, and the segmentation model is trained based on the first loss.

[0090] Of course, when determining the bone type and location of each bone in a 3D skeletal model, segmentation methods based on thresholds such as global thresholding and adaptive thresholding, edge detection, region growing and watershed methods, and graph theory can also be used to segment the image and determine the bone type and location of each bone. Since segmentation-based methods for determining the bone type and location of each bone in a 3D skeletal model are relatively mature technologies, this specification will not elaborate further.

[0091] It should be noted that the joint prosthesis described above may contain one or more bones. The number and type of bones in a joint prosthesis can be set as needed, and this manual does not impose any restrictions on this.

[0092] S104: Based on the projection result of the adjusted three-dimensional skeletal model in the full-length image and the full-length image, predict the surgical indicators after replacing the joint prosthesis, the surgical indicators being used to assess the postoperative recovery status of the target limb.

[0093] In one or more embodiments provided in this specification, the three-dimensional bone model is a three-dimensional structure, while the full-length image is a two-dimensional structure. To predict the postoperative recovery status of the target limb based on the two-dimensional and three-dimensional structures, the projection result of the three-dimensional structure in the two-dimensional structure can usually be used to perform similarity matching between the projection result and the two-dimensional structure, and the surgical indicators of the target limb after joint replacement can be predicted based on the matching result.

[0094] Based on this, the server can determine the projection result of the adjusted 3D skeletal model in the full-length image.

[0095] Specifically, the server can obtain the adjustment results of the reconstructed three-dimensional skeletal model determined in step S102 based on the expected skeletal position relationships.

[0096] Then, the server can determine the bone type and position of each bone in the adjusted 3D skeletal model, as well as the bone type and position of each bone in the full-length image, based on the adjusted 3D skeletal model.

[0097] Finally, based on the determined bone type and bone location, the adjusted 3D bone model is projected onto the full-length image.

[0098] After determining the projection result of the three-dimensional skeleton model in the full-length image, the server can predict the surgical indicators of the target limb after the joint prosthesis is replaced based on the projection result and the full-length image.

[0099] Specifically, the server can use the projection result and the full-length image as input to a pre-trained prediction model to obtain the surgical indicators output by the prediction model, which are then used as surgical indicators after replacing the joint prosthesis. These surgical indicators are used to assess the postoperative recovery status of the target limb.

[0100] Taking hip replacement surgery as an example, the surgical indicators may include leg length, the offset distance of the femoral head center relative to the acetabulum center, and prosthesis position. Taking knee replacement surgery as an example, the surgical indicators may include lower limb alignment, the angle between the femur and tibia, the offset distance between bones, and prosthesis position. The specific types of surgical indicators can be set as needed, and this instruction manual does not impose any restrictions on this.

[0101] The prediction model can be trained using the following methods:

[0102] First, acquire full-length images of several target limbs and three-dimensional skeletal models of the surgical areas of the target limbs. Each full-length image and three-dimensional skeletal model corresponds one-to-one.

[0103] Secondly, for each pair of corresponding full-length images and three-dimensional skeletal models, the three-dimensional skeletal model of the surgical area is reconstructed after replacing the bones in the three-dimensional skeletal model with joint prostheses, and the adjustment result is determined based on the expected bone positional relationship, and the reconstructed three-dimensional skeletal model is adjusted accordingly.

[0104] Then, the projection result of the adjustment result in the full-length image is determined, and the projection result and the full-length image are used as the first training sample pair, and the surgical indicators of the first training sample pair are determined as annotations.

[0105] Finally, each first training sample pair is used as input to the prediction model to be trained, and the surgical indicators corresponding to each first training sample pair output by the prediction model are obtained. Based on the surgical indicators and labels of each first training sample, the second loss is determined, and the model parameters of the prediction model are adjusted with the goal of minimizing the second loss.

[0106] The training process of the aforementioned prediction model is executed by the server that trains the model. This server and the server that executes the method for predicting surgical indicators can be the same server or different servers. After the model is trained, its parameters are stored. When the method for predicting surgical indicators needs to be executed, the model parameters are called to perform the pre-stored steps.

[0107] Of course, since the steps described above—reconstructing the 3D skeletal model of the surgical area after replacing the bones in the 3D skeletal model with joint prostheses, and adjusting the reconstructed 3D skeletal model according to the expected bone positional relationships—can all be performed by the server, the server can also directly use the full-length image of the target limb and the 3D skeletal model of the surgical area of ​​the target limb as training sample pairs to train the model in order to quickly obtain the postoperative recovery status of the target limb during application.

[0108] Specifically, the server training the model can first acquire full-length images of several target limbs and three-dimensional skeletal models of the surgical areas of the target limbs, as each second training sample pair. The annotation of each second training sample pair is then determined.

[0109] Secondly, for each second training sample pair, this second training sample pair is used as input to the reconstruction layer of the prediction model to be trained. Based on the pre-stored 3D structure of each joint prosthesis, the bones in the 3D skeletal model contained in the second training sample pair are replaced with joint prostheses, and the 3D skeletal model of the surgical area is reconstructed. The reconstructed 3D skeletal model is then adjusted according to the expected bone positional relationships.

[0110] Then, the adjustment result is used as input to the projection layer of the prediction model to determine the projection result of the adjustment result in the full-length image.

[0111] Finally, the projection result and the full-length image are used as input to the prediction layer of the prediction model to obtain the surgical indicators of the second training sample pair output by the prediction layer. Then, based on the surgical indicators and annotations of each second training sample pair, a third loss is determined, and the model parameters are adjusted based on the third loss.

[0112] Similarly, the server for training the prediction model, the server for training the forecast model, and the server for executing the method for predicting surgical indicators can be the same server or different servers.

[0113] based on Figure 1 This method for predicting surgical indicators involves acquiring a full-length image of the target limb and a 3D skeletal model of the surgical area. The 3D skeletal model is then reconstructed by replacing the bones with a joint prosthesis. Adjustments are made to the reconstructed model based on the expected skeletal relationships. Finally, based on the projection of the adjusted 3D skeletal model onto the full-length image and the full-length image itself, surgical indicators for assessing postoperative recovery of the target limb after joint replacement are predicted. This method uses a 3D skeletal model of the surgical area as a medium to project the prosthesis onto the full-length image. Based on the full-length image and the projection results, surgical indicators after joint replacement are predicted. This method eliminates the need for multi-angle X-ray imaging of the target limb, resulting in lower radiation doses and less harm to the patient.

[0114] In addition, since the model training process requires a large number of samples, if the number of samples is insufficient, it may be impossible to train an accurate prediction model or an estimated model. Therefore, this specification also provides an example in which the server directly predicts surgical indicators based on the projection results and the full-length image, so as to ensure that the surgical indicators of the target limb after joint replacement can be pre-stored even when the number of samples is small.

[0115] Specifically, the server can first perform a similarity match between the projection result and the full-length image.

[0116] Then, the server can determine the gap between the full-length image and the projection result based on the matching result, and predict surgical indicators based on the gap. Taking a knee joint prosthesis as an example, the surgical indicators include at least one of the following: lower limb alignment, recovery period and recovery degree determined based on lower limb alignment and the patient's physical condition.

[0117] Furthermore, the full-length image of the patient's target limb is typically obtained with the patient in an upright position. Compared to a 3D skeletal model that only includes the surgical area, surgical parameters predicted based on the patient's full-length image are more accurate. However, since the full-length image is determined before the joint prosthesis replacement, determining surgical parameters solely based on the full-length image results in parameters for the patient as they are now, not for after the joint prosthesis replacement. Therefore, the server can adjust the bone positions in the full-length image based on the projection of the 3D skeletal model onto the full-length image, thus accurately determining the surgical parameters based on the adjusted full-length image.

[0118] Based on this, the server can adjust the position of bones in the full-length image according to the projection results.

[0119] Specifically, the server can determine the bone type and bone position of each bone in the projection result corresponding to the joint prosthesis based on the projection result.

[0120] The server can then determine the bone location and bone type of each bone contained in the full-length image.

[0121] Finally, the server uses the center point of the joint prosthesis as the center to register the projected result with the regions of the same bone in the full-length image. Based on the registration result, the bone positions in the full-length image are adjusted. For example... Figure 4 As shown.

[0122] Figure 4 This diagram illustrates the adjustment of the full-length image provided in this specification. The left side shows the full-length image of the patient's target limb; the middle side shows a diagram of projecting the reconstructed 3D skeletal structure of the patient's surgical area onto this full-length image; and the right side shows the result of adjusting the bone positions in the full-length image based on the projection. The target limb is the lower limb. The server performs registration centered on the center point of the joint prosthesis. Figure 2 and Figure 3 Similarly, in the image, the white bone-like objects represent the patient's bones, and the gray prosthesis-like objects represent joint prostheses.

[0123] Of course, the specific center for registration can be set as needed, and this manual does not impose any restrictions on this.

[0124] After determining the adjustment results of the full-length image, the server can determine the surgical parameters after replacing the joint prosthesis based on the adjustment results.

[0125] Specifically, the server can determine the bone type and location of each bone in the adjusted full-length photographic image.

[0126] Then, the server can determine the location of the center point of each bone.

[0127] Finally, the server can determine the stress on each bone based on its bone type, location, and center point, and use this stress and center point location as surgical indicators. Figure 5 As shown.

[0128] Figure 5 This diagram illustrates the determination of surgical parameters provided in this manual. The left side of the diagram shows the full-length image of the patient's target limb obtained in step S100, and the right side shows the full-length image adjusted based on the projection results. The center point of the left knee joint is inside the lower limb force line; therefore, the image on the left corresponds to genu valgum (knee valgus). In the right image, the lower limb force line passes through the center point of the knee joint, indicating that the image on the right represents a normal knee joint. The solid black lines represent the lower limb force line, and the dashed black lines represent the midlines of each bone.

[0129] Taking the knee joint as an example, the required skeletal stress condition to be determined is the lower limb offline stress condition. This lower limb offline stress condition refers to the mechanical axis of the lower limb in the coronal plane of the knee joint and its related directional angles. Alternatively, it can be the anatomical axis in the coronal plane of the knee joint and its related directional angles, and / or the anatomical axis in the sagittal plane of the knee joint and its related directional angles. The specific type of skeletal stress condition can be set as needed, and this manual does not impose any restrictions on it.

[0130] After determining the surgical parameters for the replacement joint prosthesis, the server can use these parameters as indicators for the patient's post-articular joint replacement surgery. These parameters can then be used to assess the patient's recovery status, such as whether the stress distribution meets standards and the post-operative recovery time determined based on the amount of bone removed.

[0131] In addition, for the same patient, the matching difficulty between the 3D skeletal model of the patient's own surgical area and the full-length image of the target limb is usually the lowest, and the matching result is usually the most accurate. Therefore, in step S104, when determining the projection of the reconstructed 3D skeletal model in the full-length image, the 3D skeletal model of the patient's surgical area and the full-length image can be matched first to determine the area of ​​the 3D skeletal model in the full-length image, and the reconstructed and adjusted 3D skeletal model can be projected based on this area.

[0132] Specifically, the server can register the full-length image and the three-dimensional skeletal model of the surgical area to determine the region of the surgical area in the full-length image, which is then designated as the first region.

[0133] Then, the server can project the adjustment results of the reconstructed 3D skeletal model into the first region based on the first region.

[0134] Finally, the server can determine the image located in the first region of the full-length image, replace the image in the first region with the projection result, reconstruct the full-length image of the target limb, and adjust the position of each bone in the reconstructed full-length image according to the projection result.

[0135] Of course, when projecting the 3D skeleton model onto the first region, the server can determine the correspondence between the 3D skeleton model and the bones in the full-length image based on the bone type and bone structure of the bones contained in the first region, as well as the bone type and bone structure of the bones contained in the 3D skeleton model, and then perform projection based on the correspondence.

[0136] Furthermore, since the 3D skeletal model can be adjusted in posture, given the known acquisition posture of the full-length image of the patient's target limb, the projection result of the adjusted 3D skeletal model onto the full-length image can be quickly determined based on the posture of the 3D skeletal model and the acquisition posture of the full-length image.

[0137] Specifically, the server can adjust the pose of the 3D skeletal model to the same pose as the acquisition pose of the full-length image, and then integrate the coronal plane of the 3D skeletal structure. Since the 3D skeletal model is typically a CT scan, which includes bones and muscles, and radioactive materials attenuate differently when passing through bones and muscles, the pixel value of each pixel in the integration result can be used to characterize the attenuation at that location. Because the attenuation rates of bones and muscles differ, this attenuation can be used to determine the amount of bone and muscle present at the location corresponding to that pixel. In other words, the integration result is used to characterize whether the location corresponding to that pixel contains bone, and the amount of bone (e.g., thickness). The specific content represented by the integration result can be set as needed, and this specification does not impose any limitations on this.

[0138] Finally, the integral result is projected onto the full-length image to determine the projection result. Since the integral result can be used to characterize the amount of bone and muscle at each location, the projection result based on the integral result can determine the bone locations, etc., in the CT projection result.

[0139] Furthermore, when determining the first region, the server can also use the region where the projection result of the three-dimensional skeleton model determined based on the acquisition posture of the full-length image is located as the first region.

[0140] The registration of the full-length projection image and the 3D skeletal model can be achieved using machine learning methods.

[0141] Acquire full-length images of several patients' limbs and three-dimensional skeletal images of the surgical areas of each limb, determine several third training sample pairs, and label each third training sample pair.

[0142] Then, the server can use each third training sample pair as input to the registration model to be trained, and obtain the registration result output by the registration model.

[0143] Finally, the server can determine a fourth loss based on the gap between the annotation and registration results of each training sample pair, and train the registration model based on the fourth loss.

[0144] Of course, since model training requires a large number of samples, a traversal matching method can also be used to register the 3D skeletal model and the full-length image:

[0145] The server can determine the initial transformation relationship between the first reference frame where the 3D skeleton model is located and the second reference frame where the full-length image is located, based on the acquisition posture of the full-length image and the 3D skeleton model.

[0146] Then, the server can project the 3D skeleton model into the second reference frame according to the initial transformation relationship, and determine the initial projection result of the 3D skeleton model in the second reference frame.

[0147] Finally, the server can update the transformation relationship based on the difference between the initial projection result and the full-length image, and determine the final projection result of the 3D skeleton model in the second reference frame based on the updated transformation relationship, and take the area where the final projection result is located as the first region.

[0148] In the process of updating the transformation relationship based on the gap, the transformation relationship can be adjusted according to the gap. If the gap decreases, the transformation relationship is updated; if the gap remains unchanged or increases, the transformation relationship is updated.

[0149] Of course, mutual information can also be used to determine the updated transformation relationship.

[0150] That is, by controlling the positional and rotational relationships contained in the transformation relationship between the first and second reference frames to change within a certain range, and after each change, calculating the mutual information of the image under that parameter, traversing each changed parameter, and after the traversal is completed, taking the parameter with the largest mutual information as the updated transformation relationship.

[0151] Furthermore, because there may be a significant difference between the acquisition direction and the acquisition direction of the full-length image when determining the patient's 3D skeletal model, and because the server adjusts the posture of the 3D skeletal model based on the acquisition location, the projection result may differ significantly from the actual projection due to the different irradiation directions of the radioactive material. Therefore, a Radon transform can be performed on the 3D skeletal model to determine an accurate projection result.

[0152] Of course, since Radon transform and mutual information method are already relatively mature technologies, this manual will not elaborate on them further.

[0153] Furthermore, machine learning methods can also be used to determine the bone type and location of bones contained in the full-length image.

[0154] Specifically, the server can use the acquired full-length image as input to a pre-trained classification model to obtain the bone type and bone position of each bone in the full-length image output by the classification model.

[0155] The classification model can be trained in the following way:

[0156] First, full-length images of several limbs are acquired as second training samples, and the annotations for each second training sample are determined. The annotations for the second training samples include the bone type and location of each bone.

[0157] Then, each second training sample is used as input to the classification model to be trained, and the classification result of the second training sample is obtained from the output of the classification model.

[0158] Finally, based on the classification results and labels of each second training sample, a third loss is determined, and the classification model is trained based on the third loss.

[0159] Of course, when determining the bone type and location of each bone in the full-length image, segmentation methods based on thresholds such as global thresholding and adaptive thresholding, edge detection, region growing and watershed methods, and graph theory can also be used to segment the image and determine the bone type and location of each bone. Since methods for determining the bone type and location of each bone in a 3D skeletal model by classifying images based on segmentation are already relatively mature technologies, this specification will not elaborate on them further.

[0160] Furthermore, different patients may require different joint prostheses. Therefore, multiple joint prostheses are usually preset. In order to accurately determine the joint prosthesis that matches the patient's joint, the server can determine the corresponding surgical indicators for the patient after replacing the joint prosthesis for each group of joint prostheses, and compare the determined surgical indicators with the preset surgical indicators to determine the joint prosthesis that matches the patient.

[0161] Specifically, for each joint prosthesis, the server can determine the gap between the surgical indicators of the joint prosthesis and the preset standard indicators, and determine the matching degree between the joint prosthesis and the patient's surgical area based on the gap. Then, based on the matching degree, the joint prosthesis with the highest matching degree or the joint prosthesis with a matching degree higher than the preset matching degree threshold is selected as the joint prosthesis that matches the patient's surgical area.

[0162] Furthermore, since implanting artificial prostheses usually requires removing damaged joint surfaces, and for patients, the less bone removed, the faster the postoperative recovery and the better the outcome. Therefore, the fit of a joint prosthesis can be determined based on the amount of bone removed.

[0163] Specifically, for each joint prosthesis, the server can determine the amount of bone resection for the patient when implanting the joint prosthesis based on the three-dimensional structure of the joint prosthesis, that is, the amount of bone resection corresponding to the joint prosthesis.

[0164] Then, the server can determine the similarity of the joint prosthesis based on the difference between the surgical parameters of the joint prosthesis and the preset standard parameters. This similarity is used to characterize the similarity between the surgical parameters of the joint prosthesis and the preset standard parameters.

[0165] Finally, the server can determine the weights corresponding to the amount of osteotomy and the similarity, and based on the weights of the amount of osteotomy and the similarity, the matching degree of the joint prosthesis is determined. The weights of the amount of osteotomy and the similarity can be pre-set.

[0166] The methods for predicting surgical indicators provided in one or more embodiments of this specification are based on the same idea. This specification also provides corresponding devices for predicting surgical indicators, such as... Figure 6 As shown.

[0167] Figure 6 The device for predicting surgical parameters provided in this specification includes:

[0168] The acquisition module 200 is used to acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb.

[0169] The reconstruction module 202 is used to reconstruct the three-dimensional bone model of the surgical area after replacing the bones in the three-dimensional bone model with joint prostheses, and to adjust the reconstructed three-dimensional bone model according to the expected bone position relationship.

[0170] The prediction module 204 is used to predict surgical indicators after replacing the joint prosthesis based on the projection result of the adjusted three-dimensional skeletal model in the full-length image and the full-length image. The surgical indicators are used to evaluate the postoperative recovery status of the target limb.

[0171] Optionally, the prediction module 204 is used to project the adjusted three-dimensional skeletal model onto the full-length image, determine the projection result, adjust the bone position in the full-length image according to the projection result, and predict the surgical indicators after replacing the joint prosthesis according to the adjusted full-length image.

[0172] Optionally, the reconstruction module 202 is used to determine the expected bone position relationship corresponding to the joint prosthesis, and adjust the bone position of at least some bones corresponding to the joint prosthesis in the reconstructed three-dimensional bone model according to the expected bone position relationship, so that the position relationship of each bone in the adjusted three-dimensional bone model is the expected bone position relationship.

[0173] Optionally, the prediction module 204 is used to register the full-length image and the adjusted three-dimensional skeleton model, determine the correspondence between the adjusted three-dimensional skeleton model and the skeleton in the full-length image, and project the adjusted three-dimensional skeleton model onto the full-length image according to the correspondence to determine the projection result.

[0174] Optionally, the prediction module 204 is configured to determine the region of the surgical area in the full-length image as a first region based on the full-length image and the three-dimensional skeletal model of the surgical area, and to determine the correspondence between the adjusted three-dimensional skeletal model and the bones in the full-length image based on the bones contained in the first region in the full-length image and the bones contained in the three-dimensional skeletal model.

[0175] Optionally, the prediction module 204 is configured to determine an initial transformation relationship between a first reference frame where the three-dimensional skeleton model is located and a second reference frame where the full-length image is located, based on the acquisition posture of the full-length image and the three-dimensional skeleton model; determine an initial projection result of the three-dimensional skeleton model in the second reference frame based on the initial transformation relationship; update the transformation relationship based on the difference between the initial projection result and the full-length image; and determine the region where the final projection result of the three-dimensional skeleton model in the second reference frame is located, based on the updated transformation relationship, as the region of the surgical area in the full-length image.

[0176] Optionally, the prediction module 204 is used to determine the bone type and bone position of each bone in the adjusted full-length image, and based on the bone type and bone position of each bone, determine the lower limb force line and / or the center position of each bone in the adjusted full-length image as surgical indicators.

[0177] Optionally, the prediction module 204 is used to determine the matching degree of each group of joint prostheses based on the surgical indicators for replacing the group of joint prostheses and preset standard indicators, and to determine the joint prosthesis that matches the patient based on the matching degree of each group of joint prostheses.

[0178] Optionally, the prediction module 204 is used to determine the osteotomy amount corresponding to each joint prosthesis based on the three-dimensional structure of the joint prosthesis, determine the similarity of the joint prosthesis based on the difference between the surgical indicators of the joint prosthesis and the preset standard indicators, and determine the matching degree of the joint prosthesis based on the weight of the osteotomy amount and the preset weight of the similarity.

[0179] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The method provided is for predicting surgical indicators.

[0180] This instruction manual also provides Figure 7 The diagram shows the schematic structure of the electronic device. As described in section 7, at the hardware level, this electronic device includes a processor, an internal bus, a network interface, memory, and non-volatile memory, and may also include other hardware required for various services. The processor reads the corresponding computer program from the non-volatile memory into memory and then runs it to achieve the above-mentioned functions. Figure 1 The method for predicting surgical indicators is described above. Of course, besides software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution entity of the following processing flow is not limited to individual logic units, but can also be hardware or logic devices.

[0181] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must also be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0182] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0183] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0184] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0185] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0186] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0187] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0188] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0189] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0190] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0191] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0192] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0193] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0194] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0195] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0196] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for predicting surgical indicators, characterized in that, include: Acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb; After replacing the bones in the three-dimensional skeletal model with joint prostheses, the three-dimensional skeletal model of the surgical area is reconstructed, and the reconstructed three-dimensional skeletal model is adjusted according to the expected bone positional relationship. Based on the projection of the adjusted three-dimensional skeletal model onto the full-length image and the full-length image, surgical indicators are predicted after the replacement of the joint prosthesis. These surgical indicators are used to assess the postoperative recovery status of the target limb.

2. The method as described in claim 1, characterized in that, Based on the projection of the adjusted 3D skeletal model onto the full-length image and the full-length image itself, surgical parameters after joint prosthesis replacement are predicted, specifically including: The adjusted 3D skeletal model is projected onto the full-length image, and the projection result is determined; Based on the projection results, the positions of the bones in the full-length image are adjusted; Based on the adjusted full-length image, surgical parameters after joint prosthesis replacement are predicted.

3. The method as described in claim 1, characterized in that, Based on the expected skeletal positional relationships, the reconstructed 3D skeletal model is adjusted, specifically including: Determine the expected skeletal positional relationship corresponding to the joint prosthesis; Based on the expected bone position relationship, the bone positions of at least some of the bones corresponding to the joint prosthesis in the reconstructed three-dimensional bone model are adjusted so that the position relationship of each bone in the adjusted three-dimensional bone model is the expected bone position relationship.

4. The method as described in claim 1, characterized in that, Determining the projection result of the adjusted 3D skeletal model into the full-length image specifically includes: The full-length image and the adjusted 3D skeletal model are registered to determine the correspondence between the bones in the adjusted 3D skeletal model and the full-length image. Based on the correspondence, the adjusted 3D skeletal model is projected onto the full-length image to determine the projection result.

5. The method as described in claim 4, characterized in that, Determining the correspondence between the adjusted 3D skeletal model and the bones in the full-length image specifically includes: Based on the full-length image and the three-dimensional skeletal model of the surgical area, the region of the surgical area in the full-length image is determined as the first region; Based on the bones contained in the first region of the full-length image and the bones contained in the three-dimensional bone model, the correspondence between the adjusted three-dimensional bone model and the bones in the full-length image is determined.

6. The method as described in claim 5, characterized in that, Based on the full-length image and the three-dimensional skeletal model of the surgical area, the region of the surgical area within the full-length image is determined, specifically including: Based on the acquisition posture of the full-length image and the three-dimensional skeleton model, determine the initial transformation relationship between the first reference frame where the three-dimensional skeleton model is located and the second reference frame where the full-length image is located; Based on the initial transformation relationship, determine the initial projection result of the three-dimensional skeleton model in the second reference frame; Based on the difference between the initial projection result and the full-length image, the transformation relationship is updated, and based on the updated transformation relationship, the region where the final projection result of the three-dimensional skeleton model is located in the second reference frame is determined as the region of the surgical area in the full-length image.

7. The method as described in claim 2, characterized in that, Based on the adjusted full-length image, predict surgical parameters after joint prosthesis replacement, specifically including: Determine the bone type and bone position of each bone in the adjusted full-length image; Based on the bone type and location of each bone, the lower limb force line and / or the center position of each bone in the adjusted full-length image are determined as surgical indicators.

8. The method as described in claim 1, characterized in that, The joint prosthesis has multiple sets, and the method further includes: For each group of joint prostheses, the matching degree of the group of joint prostheses is determined based on the surgical indicators for replacing the group of joint prostheses and the preset standard indicators; Based on the matching degree of each group of joint prostheses, a joint prosthesis that matches the patient is determined.

9. The method as described in claim 8, characterized in that, Based on the surgical parameters and pre-set standard parameters for replacing this group of joint prostheses, the matching degree of this group of joint prostheses was determined, specifically including: Based on the three-dimensional structure of this group of joint prostheses, the corresponding osteotomy amount for this group of joint prostheses is determined; The similarity of the joint prostheses in this group is determined based on the difference between the surgical indicators and the preset standard indicators. The matching degree of the joint prosthesis group is determined based on the osteotomy amount and the preset weight of the osteotomy amount, as well as the similarity and the preset weight of the similarity.

10. A device for predicting surgical indicators, characterized in that, The device includes: The acquisition module is used to acquire a full-length image of the patient's target limb and a three-dimensional skeletal model of the surgical area of ​​the target limb; The reconstruction module is used to reconstruct the three-dimensional bone model of the surgical area after replacing the bones in the three-dimensional bone model with joint prostheses, and to adjust the reconstructed three-dimensional bone model according to the expected bone position relationship. The prediction module is used to predict surgical indicators after replacing the joint prosthesis based on the projection result of the adjusted three-dimensional skeletal model in the full-length image and the full-length image. The surgical indicators are used to assess the postoperative recovery status of the target limb.

11. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 9.

12. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 9.

Images