Affected part recovery prediction method, device, equipment and medium

Abstract

The embodiments of the present application relate to the field of medical diagnosis and disclose a method, apparatus, equipment and medium for predicting the rehabilitation of an affected part, the method comprising: obtaining scanning data of a patient's affected part and extracting features of the affected part; reconstructing a human affected part model based on the features of the affected part; generating a state prediction map of the affected part in each stage of the rehabilitation process based on the prediction of the human affected part model; obtaining the patient's treatment information to generate a correction constraint corresponding to the state prediction map based on the treatment information; correcting the state prediction map based on the correction constraint to obtain a target prediction map; assigning a time coordinate to the target prediction map based on the treatment information, the time coordinate being the expected time corresponding to the affected part when it recovers to a stage matching the target prediction map; associating the target prediction map with the time axis based on the time coordinate to generate a rehabilitation prediction atlas, so as to accurately predict the rehabilitation progress of the patient's affected part, and estimate the rehabilitation effect in advance based on the treatment plan, and clearly and intuitively display the evolution of the affected part in subsequent treatment.

Description

Technical Field

[0001] The present application relates to the field of medical diagnosis, and in particular to a method, device, equipment and medium for predicting the recovery of an affected area. Background Art

[0002] Imaging refers to the technology and process of obtaining non-invasive images of the human body or a part of it. In the medical field, imaging can be used to determine the condition of the affected area, either subcutaneously or externally. For example, imaging scans such as MRI, CT, and X-rays can be used to obtain the actual condition of the affected area.

[0003] However, at present, simply showing the results of imaging examinations to doctors and patients cannot clearly and intuitively reflect the evolution of the affected area during subsequent treatment. This makes it inconvenient for doctors and patients to intuitively understand and identify the symptoms and treatment plans of the disease, and it is even more inconvenient for patients to know the effect of rehabilitation of the affected area under the current treatment conditions. Summary of the Invention

[0004] The main purpose of the embodiments of the present application is to provide a method, device, equipment and medium for predicting the rehabilitation of an affected part, aiming to accurately predict the rehabilitation progress of the patient's affected part and estimate the rehabilitation effect in advance based on the treatment plan, so as to clearly and intuitively show the physician and patient the evolution of the affected part in subsequent treatment.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting recovery of an affected area, comprising:

[0006] obtaining scan data obtained by scanning an affected part of a patient, and analyzing the scan data to extract affected part features of the affected part;

[0007] Reconstruct the human affected area model according to the characteristics of the affected area;

[0008] Generate a prediction map of the affected part's status at each stage of the rehabilitation process based on the human affected part model;

[0009] Obtain the patient's treatment information and generate correction constraints corresponding to the state prediction graph based on the treatment information;

[0010] Performing prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph;

[0011] Assigning a time coordinate to each target prediction map based on the treatment information, where the time coordinate is the expected time corresponding to when the affected part recovers to the stage matching the target prediction map;

[0012] The target prediction map is associated with the time axis according to the corresponding time coordinate to generate a rehabilitation prediction atlas of the affected part.

[0013] In some embodiments, the correction constraint includes at least a recovery degree constraint and a recovery speed constraint of the affected part;

[0014] The state prediction graph is predicted and corrected according to the correction constraints to generate the corresponding target prediction graph, including:

[0015] Extracting lesion features from the state prediction graph;

[0016] The lesion features are modified according to the recovery degree constraint and the recovery speed constraint to obtain the modified features;

[0017] Reconstruct and generate a state correction map based on the correction feature;

[0018] The state correction graph is filtered according to the recovery degree constraint to obtain the target prediction graph.

[0019] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. Generating a modified constraint corresponding to the state prediction graph based on the treatment information includes:

[0020] Generate recovery level constraints based on health status;

[0021] Generate recovery speed constraints based on health status and treatment plan.

[0022] In some embodiments, the lesion feature includes a location feature and a first contour feature:

[0023] The lesion features are modified according to the recovery degree constraint and the recovery speed constraint to obtain the modified features, including:

[0024] Generate a regression coefficient based on the recovery speed constraint and the position characteristics;

[0025] Converting the first contour feature into a first area feature value and a first morphological feature value;

[0026] A second contour feature is generated according to the fading coefficient, the first area feature value, and the first morphological feature value to obtain a modified feature including a position feature and a second contour feature.

[0027] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. Assigning a time coordinate to each target prediction graph based on the treatment information includes:

[0028] Determine the target degree of recovery based on the health status and the target total period required to achieve the target degree of recovery under the conditions of the health status and treatment plan;

[0029] Determine the sub-cycles required to achieve the degree of recovery corresponding to the target prediction graph based on the treatment plan;

[0030] The time coordinate is assigned to the corresponding target prediction map according to the relative proportion of the sub-cycle to the target total cycle.

[0031] In some embodiments, analyzing the scan data to extract features of the affected area includes:

[0032] generating first diagnosis and treatment information according to the scan data;

[0033] Obtaining the diagnosis and treatment evaluation input by the physician, and performing semantic recognition on the diagnosis and treatment evaluation to generate second diagnosis and treatment information;

[0034] An affected area feature is generated according to the first diagnosis and treatment feature and the second diagnosis and treatment feature.

[0035] In some embodiments, reconstructing a human affected area model based on affected area characteristics includes:

[0036] Determine the corresponding affected area according to the characteristics of the affected area;

[0037] An affected part reference model corresponding to the affected part is called, and parameters of the affected part reference model are adjusted according to the affected part characteristics to generate a human affected part model.

[0038] In a second aspect, an embodiment of the present application further provides an affected part recovery prediction device, comprising:

[0039] a feature extraction module for acquiring scan data obtained by scanning the affected part of the patient and analyzing the scan data to extract affected part features of the affected part;

[0040] A model reconstruction module is used to reconstruct a human affected area model based on the characteristics of the affected area;

[0041] A state prediction module is used to generate a state prediction map of the affected part at each stage of the rehabilitation process based on the human affected part model;

[0042] A constraint generation module is used to obtain the patient's treatment information and generate correction constraints corresponding to the state prediction graph based on the treatment information;

[0043] A prediction correction module, configured to perform prediction correction on the state prediction graph according to the correction constraints to generate a corresponding target prediction graph;

[0044] A time prediction module is used to assign a time coordinate to each target prediction map based on the treatment information, wherein the time coordinate is the expected time corresponding to when the affected part recovers to a stage matching the target prediction map;

[0045] The atlas generation module is used to associate the target prediction map with the time axis according to the corresponding time coordinates to generate a rehabilitation prediction atlas of the affected part.

[0046] In a third aspect, an embodiment of the present application further provides a computer device, the computer device including a memory and a processor;

[0047] Memory for storing computer programs;

[0048] A processor is used to execute a computer program and implement the steps of any one of the methods for predicting the recovery of an affected area provided in the embodiments of the present application when executing the computer program.

[0049] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the steps of any of the affected area rehabilitation prediction methods provided in the embodiments of the present application.

[0050] In summary, the embodiments of the present application provide a method, apparatus, device and medium for predicting the rehabilitation of an affected part, wherein the method includes: obtaining scanning data obtained by scanning the affected part of a patient, and analyzing the scanning data to extract the affected part features of the affected part; reconstructing a human affected part model according to the affected part features; predicting and generating a state prediction map of the affected part in each stage of the rehabilitation process based on the human affected part model; obtaining the patient's treatment information, and generating correction constraints corresponding to the state prediction map according to the treatment information; predicting and correcting the state prediction map according to the correction constraints to generate a corresponding target prediction map; assigning a time coordinate to each target prediction map according to the treatment information, wherein the time coordinate is the expected time corresponding to the affected part when it recovers to a stage matching the target prediction map; associating the target prediction map with the time axis according to the corresponding time coordinate to generate a rehabilitation prediction map set of the affected part, so as to accurately predict the rehabilitation progress of the patient's affected part, and estimate the rehabilitation effect in advance according to the treatment plan, so as to clearly and intuitively show the physician and the patient the evolution of the affected part in subsequent treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0052] Figure 1 A schematic diagram of a flow chart of a method for predicting the recovery of an affected area provided in an embodiment of the present application;

[0053] Figure 2 A schematic diagram of a process for generating a target prediction map in a method for predicting the recovery of an affected area provided in an embodiment of the present application;

[0054] Figure 3A schematic diagram of generating a target prediction map in an affected area rehabilitation prediction method provided in an embodiment of the present application;

[0055] Figure 4 For Figure 2 Provided is a schematic diagram of the steps involved in obtaining correction features during the process of generating a target prediction map;

[0056] Figure 5 A schematic diagram of a rehabilitation prediction atlas generated by an affected part rehabilitation prediction method provided in an embodiment of the present application;

[0057] Figure 6 A schematic block diagram of the structure of an affected part recovery prediction device provided in an embodiment of the present application;

[0058] Figure 7 A schematic block diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

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

[0060] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0061] It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0062] Imaging refers to the technology and process of obtaining non-invasive images of the human body or a part of it. In the medical field, imaging can be used to determine the condition of the affected area, either subcutaneously or externally. For example, imaging scans such as MRI, CT, and X-rays can be used to obtain the actual condition of the affected area.

[0063] However, at present, simply showing the results of imaging examinations to doctors and patients cannot clearly and intuitively reflect the evolution of the affected area during subsequent treatment. This makes it inconvenient for doctors and patients to intuitively understand and identify the symptoms and treatment plans of the disease, and it is even more inconvenient for patients to know the effect of rehabilitation of the affected area under the current treatment conditions.

[0064] To address the above issues, embodiments of the present application provide a method, apparatus, device, and medium for predicting the recovery of an affected area. The following describes some embodiments of the present application in detail, with reference to the accompanying drawings. The following embodiments and features thereof may be combined with one another unless they conflict.

[0065] It should be noted that the affected area recovery prediction method can be applied to a computer device, such as a mobile phone, tablet computer, laptop computer, desktop computer, personal digital assistant, wearable device, etc., or a server terminal, wherein the server terminal can be a standalone server or a server cluster. In the embodiments of this application, the method is specifically described using a standalone server as an example.

[0066] Please refer to Figure 1 , Figure 1 A flowchart of a method for predicting the recovery of an affected area provided in an embodiment of the present application.

[0067] like Figure 1 As shown, the affected part recovery prediction method includes steps S1 to S7.

[0068] Step S1: obtaining scan data obtained by scanning the affected part of the patient, and analyzing the scan data to extract affected part features of the affected part.

[0069] Specifically, scan data refers to information obtained by scanning a patient's affected area, such as the shadow depth, shadow contour, and diseased cell concentration of the affected area. Scanning a patient's affected area, for example, involves performing an imaging examination of the affected area, specifically obtaining images of the internal tissues of the human body or a portion of the human body through imaging examination in a non-invasive manner. For further example, image data from medical images such as MRI (Nuclear Magnetic Resonance Imaging), CT (Computed Tomography), and X-rays can be obtained through imaging scanning, and such image data can be used as scan data.

[0070] After acquiring the scan data, feature analysis is performed on the scan data to extract the affected area features. For example, the scan data can be input into a preset medical image analysis model, so that the medical image analysis model extracts and outputs the affected area features, such as the affected area shadow depth, the affected area shadow contour, the diseased cell concentration and other features.

[0071] In some embodiments, analyzing the scan data to extract features of the affected area includes:

[0072] generating first diagnosis and treatment information according to the scan data;

[0073] Obtaining the diagnosis and treatment evaluation input by the physician, and performing semantic recognition on the diagnosis and treatment evaluation to generate second diagnosis and treatment information;

[0074] An affected area feature is generated according to the first diagnosis and treatment feature and the second diagnosis and treatment feature.

[0075] It should be noted that the first diagnosis and treatment information is derived from the scan data, such as the affected area shadow depth, affected area shadow outline, and diseased cell concentration, while the second diagnosis and treatment information is derived from semantic recognition based on the diagnosis and treatment evaluation input by the physician. When generating the affected area features, the first diagnosis and treatment information derived from the scan data and the second diagnosis and treatment information derived from the physician's diagnosis and treatment evaluation are first obtained, and then the affected area features of the corresponding affected area are determined based on these two information.

[0076] For example, the affected part is the patient's lungs. An MRI scan of the patient's affected part can obtain an MRI image as scan data. When extracting the characteristics of the affected part of the patient's lungs, the server executing this method can obtain characteristics such as lung shadow depth, lung shadow contour, and lung diseased cell concentration based on the MRI image as first diagnosis and treatment information; in addition, the physician can input the diagnosis and treatment information of the affected part into this server as second diagnosis and treatment information based on the MRI image and the patient's previous treatment, and the server then generates characteristics of the affected part of the lungs based on the first diagnosis and treatment characteristics and the second diagnosis and treatment characteristics.

[0077] Step S2: Reconstructing a human affected area model according to the affected area characteristics.

[0078] The human affected part model may be a three-dimensional model or a plurality of two-dimensional images, and the human affected part model is used to show the basic condition of the patient's affected part.

[0079] In some embodiments, reconstructing a human affected area model based on affected area characteristics includes:

[0080] Determine the corresponding affected area according to the characteristics of the affected area;

[0081] An affected part reference model corresponding to the affected part is called, and parameters of the affected part reference model are adjusted according to the affected part characteristics to generate a human affected part model.

[0082] Specifically, the affected part reference model is a three-dimensional model with preset adjustable parameters, which corresponds to the morphology of the affected part. When reconstructing the human affected part model based on the characteristics of the affected part: first, the corresponding affected part is identified based on the characteristics of the affected part, such as the lungs, knees, inside the mouth or other parts of the human body organs, and then the affected part reference model corresponding to the affected part is called, and the parameters of the affected part reference model are adjusted according to the characteristics of the affected part to generate the human affected part model. For example, when the affected part is the patient's lung, the corresponding lung reference model is called as the affected part reference model, and then according to the characteristic numerical values ​​of the affected part characteristics "lung shadow depth" and "lung shadow contour", the parameters are adjusted in the affected part reference model to arrange the shadow that matches the morphology in the lung reference model, and the human affected part model corresponding to the patient's lung is obtained.

[0083] It should be understood that this solution adjusts the parameters of the affected area reference model to generate a human affected area model. Without prior input of affected area information, this solution can automatically determine the specific affected area based solely on the input affected area features, intelligently call the corresponding affected area reference model, and thus generate a human affected area model. For example, by simply inputting the affected area features "lung shadow depth" and "lung shadow outline" to this server, the server can determine the corresponding affected area as the lung based on these affected area features, and then call the corresponding lung reference model to generate the human affected area model, greatly improving the intelligence of generating human affected area models.

[0084] Step S3: Generate a state prediction diagram of the affected part in each stage of the rehabilitation process based on the human affected part model prediction.

[0085] It should be noted that the human affected part model can be a three-dimensional model or multiple two-dimensional images, which are used to represent and display the current basic conditions of the patient's affected part, especially the current morphology of the affected part.

[0086] Specifically, when generating the state prediction map, a preset image generation model is first invoked. The human affected part model is then input into the image generation model, allowing the image generation model to predict and generate state prediction maps of the affected part at each stage of the rehabilitation process based on the current basic condition of the affected part. The state prediction map of the affected part is a prediction of the subsequent morphology of the affected part during the rehabilitation process, based on the current morphology of the affected part represented in the human affected part model.

[0087] For example, image generation models such as DALL-E2 or GPT (Generative Pre-training Transformer) can be used. Taking DALL-E2 as an example, DALL-E2 is a machine learning-based text-to-image AI art generator. By inputting a model of a human affected area into DALL-E2, DALL-E2 can identify the semantics corresponding to the model of the affected area and predict the state of the affected area at each stage of the subsequent recovery process.

[0088] Step S4: Obtain the patient's treatment information, and generate correction constraints corresponding to the state prediction graph based on the treatment information.

[0089] It should be noted that for the same current affected area morphology, based on the treatment information of different patients, the rehabilitation changes, rehabilitation trends, rehabilitation speed and rehabilitation degree of the affected area in the subsequent rehabilitation process will be different, and the correction constraint is used to characterize the actual adjustment direction of the state prediction diagram under the influence of the treatment information, wherein the correction constraint can be a set of parameter factors, and introducing at least part of the parameters of the state prediction diagram into the parameter factor can perform prediction correction to generate the corresponding target prediction diagram.

[0090] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. Generating a modified constraint corresponding to the state prediction graph based on the treatment information includes:

[0091] Generate recovery level constraints based on health status;

[0092] Generate recovery speed constraints based on health status and treatment plan.

[0093] It should be noted that treatment information includes at least the patient's health status and the treatment plan adopted by the patient. The patient's health status may include information such as the patient's age and medical history. For the same affected area, the subsequent recovery process, recovery trends, recovery speed, and degree of recovery may vary depending on the health status and treatment plan.

[0094] Furthermore, the correction constraints corresponding to the state prediction diagram are generated based on the treatment information, specifically based on the patient's own different health conditions and the treatment plan adopted for the patient, wherein the correction constraints at least include a recovery degree constraint and a recovery speed constraint.

[0095] It should also be noted that the recovery degree constraint represents the degree of recovery that can be achieved under the current health status and treatment plan. The recovery degree constraint is strongly correlated with the health status, so a recovery degree constraint can be generated based on the health status. The recovery rate constraint, on the other hand, represents the speed at which the affected part can recover under the current health status and treatment plan. The recovery degree constraint is strongly correlated with both the health status and the treatment plan, so a recovery degree constraint can be generated based on the health status and the treatment plan.

[0096] Exemplarily, the rehabilitation degree constraint and / or the rehabilitation speed constraint may be parameter factors, and introducing at least part of the parameters of the state prediction map into the parameter factors may perform prediction correction to generate a corresponding target prediction map.

[0097] Step S5: Perform prediction correction on the state prediction graph according to the correction constraints to generate a corresponding target prediction graph.

[0098] After obtaining the correction constraints, including the recovery degree constraint and the recovery speed constraint, the state prediction graph is subjected to a prediction correction based on the correction constraints to generate a corresponding target prediction graph. For example, if the correction constraints are in the form of a set of parameter factors, at least some of the parameters of the state prediction graph can be introduced into the parameter factors to perform a prediction correction to generate a corresponding target prediction graph.

[0099] It should be noted that the target prediction graph represents: after introducing treatment information including the patient's health status and the treatment plan adopted by the patient, the prediction process is corrected to take into account the impact of the treatment information on the state prediction graph.

[0100] like Figure 2 and Figure 3 As shown, in some embodiments, the correction constraint includes at least a recovery degree constraint and a recovery speed constraint of the affected part, and step S5 performs a prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph, including steps S51-S54:

[0101] Step S51: extracting lesion features in the state prediction graph;

[0102] Step S52: Correcting the lesion feature according to the recovery degree constraint and the recovery speed constraint to obtain a corrected feature;

[0103] Step S53: reconstructing and generating a state correction map based on the correction feature;

[0104] Step S54: Screening the state correction graph according to the recovery degree constraint to obtain a target prediction graph.

[0105] It should be noted that a lesion refers to a localized area of ​​diseased tissue in the body that harbors pathogenic microorganisms. For example, if a part of the lung is damaged by tuberculosis bacteria, this part is a pulmonary tuberculosis lesion. For example, when the lung is the affected area, the shadowed area in the lung MRI image can be regarded as the lesion.

[0106] It should also be noted that lesion features refer to the external morphological features of the lesion. For example, when the lungs are the affected area, the lung MRI image, the human affected area model corresponding to the lungs, and the shadow depth and shadow outline of the shadow part in the state prediction diagram can all be used as lesion features.

[0107] Specifically, the device executing this method first extracts the lesion features in the state prediction map, then corrects the lesion features according to the recovery degree constraint and the recovery speed constraint to obtain the corrected features, and reconstructs the state correction map based on the corrected features, and then screens the state correction map according to the recovery degree constraint to obtain the target prediction map.

[0108] It should be understood that due to the patient's health condition and the treatment plan adopted by the patient, the degree of recovery of the patient's affected part may not be able to fully reach the diseased state. Therefore, this solution then filters out the pictures whose corresponding states cannot be reached in multiple state correction maps based on the recovery degree constraint, and uses the remaining state correction maps as target prediction maps, thereby improving the intelligence of generating the target prediction map and the degree of consistency with the patient's actual condition.

[0109] In some embodiments, the lesion feature includes a position feature and a first contour feature. It should be noted that the position feature represents the relative positional relationship between the lesion and the affected area, while the first contour feature represents the contour morphology of the lesion in the state prediction diagram.

[0110] like Figures 3 and 4 As shown, in step S52, the lesion feature is corrected according to the recovery degree constraint and the recovery speed constraint to obtain the corrected feature, including steps S521 to S523:

[0111] Step S521: generating a regression coefficient according to the recovery speed constraint and the position feature;

[0112] Step S522: converting the first contour feature into a first area feature value and a first morphological feature value;

[0113] Step S523: generating a second contour feature according to the fading coefficient, the first area feature value, and the first morphological feature value to obtain a modified feature including a position feature and a second contour feature.

[0114] It should be noted that the second contour feature represents the contour morphology of the lesion in the state correction diagram, while the regression coefficient represents the regression trend and regression speed of the lesion contour under the constraint of the recovery speed. It should also be noted that the regression trend and regression speed of the lesion contour are also related to the relative position relationship between the lesion and the affected area, that is, the position feature.

[0115] Specifically, the server first generates a regression coefficient based on the recovery rate constraint and the position feature. It then converts the first contour feature into a first area feature value and a first morphological feature value. The first area feature value represents the area of ​​the lesion contour, and the first morphological feature value represents the morphology of the lesion contour. Subsequently, the first area feature value and the first morphological feature value are modified based on the regression coefficient to generate a second area feature value and a second morphological feature value. A second contour feature is then generated based on the second area feature value and the second morphological feature value, resulting in a modified feature that includes the position feature and the second contour feature.

[0116] Exemplarily, the first area characteristic value and the first morphological characteristic value are corrected according to the fading coefficient, specifically, the first area characteristic value and the first morphological characteristic value are respectively multiplied by the fading coefficient to generate the second area characteristic value and the second morphological characteristic value. When the fading speed of the lesion outline reaches a maximum value, the fading coefficient is 1; when the lesion outline does not reach a maximum value, the fading coefficient is less than 1.

[0117] Step S6: assigning a time coordinate to each target prediction map according to the treatment information, wherein the time coordinate is the expected time corresponding to when the affected part recovers to a stage matching the target prediction map.

[0118] Specifically, after generating multiple target prediction maps, the server assigns a time coordinate to each target prediction map based on the treatment information. The time coordinate represents the expected time for the affected area to recover to a stage matching the target prediction map. It should be understood that treatment information may include, for example, the patient's health status and the treatment plan adopted by the patient. Depending on the patient's health status and treatment plan, the expected time required for the affected area to reach the same stage may vary.

[0119] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. Assigning a time coordinate to each target prediction graph based on the treatment information includes:

[0120] Determine the target degree of recovery based on the health status and the target total period required to achieve the target degree of recovery under the conditions of the health status and treatment plan;

[0121] Determine the sub-cycles required to achieve the degree of recovery corresponding to the target prediction graph based on the treatment plan;

[0122] The time coordinate is assigned to the corresponding target prediction map according to the relative proportion of the sub-cycle to the target total cycle.

[0123] Specifically, the target recovery degree is the maximum recovery degree that the patient can achieve after treatment. The target recovery degrees that can be achieved by individual patients with different health conditions are also different. Therefore, the target recovery degree is first determined based on the health condition, and then the target total cycle required to achieve the target recovery degree under the conditions of the health condition and the treatment plan is determined. Then, based on the treatment plan, the sub-cycle required to achieve the recovery degree corresponding to the target prediction graph is determined, and the time coordinate is assigned to the corresponding target prediction graph according to the relative proportional relationship between the sub-cycle and the target total cycle.

[0124] It should be noted that the time coordinate is assigned to the corresponding target prediction graph according to the relative proportional relationship between the sub-cycle and the target total cycle, so that the time coordinate of the target prediction graph not only reflects the time required to reach the corresponding degree of recovery, but also reflects the stage of the recovery degree corresponding to any target prediction graph in the entire rehabilitation process.

[0125] By assigning time coordinates to the corresponding target prediction map based on the patient's health status and the treatment plan adopted by the patient, we can more accurately determine the actual time required for the patient's affected part to reach the recovery level corresponding to the target prediction map and the relative time in the entire rehabilitation process, making the prediction of the recovery process of the affected part more accurate and more in line with the patient's actual situation.

[0126] Step S7: Associating the target prediction graph with the time axis according to the corresponding time coordinate to generate a rehabilitation prediction atlas of the affected part.

[0127] Each target prediction graph has a corresponding time coordinate. This time coordinate not only reflects the time required to reach the corresponding recovery level, but also reflects the stage of recovery corresponding to each target prediction graph within the entire rehabilitation process. By linking the target prediction graphs to the same timeline based on their corresponding time coordinates, a set of rehabilitation prediction graphs reflecting the rehabilitation process of the affected area is obtained.

[0128] It should be noted that the obtained rehabilitation prediction atlas includes multiple target prediction maps and a time axis, and each target prediction map is associated with a corresponding coordinate on the time axis.

[0129] As Figure 5The rehabilitation prediction atlas shown in the figure is illustrated. The rehabilitation prediction atlas includes at least target prediction maps P1, P2, P3, P4, and P5. The time coordinates corresponding to the above five target prediction maps are (t1, 0), (t2, 0), (t3, 0), (t4, 0), and (t5, 0), respectively. The target prediction maps are associated with the time axis according to the corresponding time coordinates to obtain a rehabilitation prediction atlas. In the obtained rehabilitation prediction atlas, the coordinates of the target prediction maps P1, P2, P3, P4, and P5 are sorted in chronological order, so as to accurately predict the rehabilitation progress of the patient's affected part and estimate the rehabilitation effect in advance according to the treatment plan, so as to clearly and intuitively show the physician and patient the evolution of the affected part in subsequent treatment.

[0130] In summary, the embodiments of the present application provide a method, apparatus, device and medium for predicting the rehabilitation of an affected part, wherein the method includes: obtaining scanning data obtained by scanning the affected part of a patient, and analyzing the scanning data to extract the affected part features of the affected part; reconstructing a human affected part model according to the affected part features; predicting and generating a state prediction map of the affected part in each stage of the rehabilitation process based on the human affected part model; obtaining the patient's treatment information, and generating correction constraints corresponding to the state prediction map according to the treatment information; predicting and correcting the state prediction map according to the correction constraints to generate a corresponding target prediction map; assigning a time coordinate to each target prediction map according to the treatment information, wherein the time coordinate is the expected time corresponding to the affected part when it recovers to a stage matching the target prediction map; associating the target prediction map with the time axis according to the corresponding time coordinate to generate a rehabilitation prediction map set of the affected part, so as to accurately predict the rehabilitation progress of the patient's affected part, and estimate the rehabilitation effect in advance according to the treatment plan, so as to clearly and intuitively show the physician and the patient the evolution of the affected part in subsequent treatment.

[0131] The present application also provides an affected part recovery prediction device. Figure 6 , Figure 6 This is a schematic block diagram of the structure of the affected area recovery prediction device provided in an embodiment of the present application.

[0132] like Figure 6 As shown, the affected part recovery prediction device 800 includes:

[0133] The feature extraction module 801 is used to obtain scan data obtained by scanning the affected part of the patient and analyze the scan data to extract the affected part features;

[0134] The model reconstruction module 802 is used to reconstruct the human body affected part model according to the affected part characteristics;

[0135] The state prediction module 803 is used to generate a state prediction map of the affected part at each stage of the rehabilitation process based on the human affected part model;

[0136] The constraint generation module 804 is used to obtain the patient's treatment information and generate the correction constraints corresponding to the state prediction graph based on the treatment information;

[0137] The prediction correction module 805 is used to perform prediction correction on the state prediction graph according to the correction constraints to generate a corresponding target prediction graph;

[0138] A time prediction module 806 is configured to assign a time coordinate to each target prediction map based on the treatment information, wherein the time coordinate is the expected time corresponding to when the affected part recovers to a stage matching the target prediction map;

[0139] The atlas generation module 807 is used to associate the target prediction map with the time axis according to the corresponding time coordinates to generate a rehabilitation prediction atlas of the affected part.

[0140] In some embodiments, the correction constraint includes at least a recovery degree constraint and a recovery speed constraint of the affected part;

[0141] When the state prediction module 803 performs prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph, it includes:

[0142] Extracting lesion features from the state prediction graph;

[0143] The lesion features are modified according to the recovery degree constraint and the recovery speed constraint to obtain the modified features;

[0144] Reconstruct and generate a state correction map based on the correction feature;

[0145] The state correction graph is filtered according to the recovery degree constraint to obtain the target prediction graph.

[0146] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. When the constraint generation module 804 generates the modified constraints corresponding to the state prediction graph based on the treatment information, it includes:

[0147] Generate recovery level constraints based on health status;

[0148] Generate recovery speed constraints based on health status and treatment plan.

[0149] In some embodiments, the lesion feature includes a location feature and a first contour feature:

[0150] When the constraint generation module 804 modifies the lesion feature according to the rehabilitation degree constraint and the rehabilitation speed constraint to obtain the modified feature, it includes:

[0151] Generate a regression coefficient based on the recovery speed constraint and the position characteristics;

[0152] Converting the first contour feature into a first area feature value and a first morphological feature value;

[0153] A second contour feature is generated according to the fading coefficient, the first area feature value, and the first morphological feature value to obtain a modified feature including a position feature and a second contour feature.

[0154] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. When the time prediction module 806 assigns a time coordinate to each target prediction graph based on the treatment information, it includes:

[0155] Determine the target degree of recovery based on the health status and the target total period required to achieve the target degree of recovery under the conditions of the health status and treatment plan;

[0156] Determine the sub-cycles required to achieve the degree of recovery corresponding to the target prediction graph based on the treatment plan;

[0157] The time coordinate is assigned to the corresponding target prediction map according to the relative proportion of the sub-cycle to the target total cycle.

[0158] In some embodiments, when analyzing the scan data to extract the affected area features, the feature extraction module 801 includes:

[0159] generating first diagnosis and treatment information according to the scan data;

[0160] Obtaining the diagnosis and treatment evaluation input by the physician, and performing semantic recognition on the diagnosis and treatment evaluation to generate second diagnosis and treatment information;

[0161] An affected area feature is generated according to the first diagnosis and treatment feature and the second diagnosis and treatment feature.

[0162] In some embodiments, when the model reconstruction module 802 reconstructs the human affected area model according to the affected area characteristics, it includes:

[0163] Determine the corresponding affected area according to the characteristics of the affected area;

[0164] An affected part reference model corresponding to the affected part is called, and parameters of the affected part reference model are adjusted according to the affected part characteristics to generate a human affected part model.

[0165] The present application also provides a computer device. Figure 6 , Figure 6 A schematic block diagram of the structure of a computer device provided in an embodiment of the present application.

[0166] like Figure 7As shown, an embodiment of the present application further provides a computer device 900, which includes a processor 901 and a memory 902. The processor 901 and the memory 902 are connected via a bus 903, which is, for example, an I2C (Inter-integrated Circuit) bus.

[0167] Specifically, the processor 901 is used to provide computing and control capabilities to support the operation of the entire computer device. The processor 301 can be a central processing unit (CPU), and the processor 301 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0168] Specifically, the memory 902 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0169] Those skilled in the art will understand that Figure 6 The structure shown in the figure is merely a block diagram of a portion of the structure related to the embodiment of the present application, and does not constitute a limitation on the computer device to which the embodiment of the present application is applied. Specifically, the computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0170] The processor 901 is configured to execute a computer program stored in the memory, and implement any one of the methods for predicting the recovery of an affected area provided in the embodiments of the present application when executing the computer program.

[0171] In some embodiments, the processor 901 is configured to run a computer program stored in the memory, and implement the following steps when executing the computer program:

[0172] obtaining scan data obtained by scanning an affected part of a patient, and analyzing the scan data to extract affected part features of the affected part;

[0173] Reconstruct the human affected area model according to the characteristics of the affected area;

[0174] Generate a prediction map of the affected part's status at each stage of the rehabilitation process based on the human affected part model;

[0175] Obtain the patient's treatment information and generate correction constraints corresponding to the state prediction graph based on the treatment information;

[0176] Performing prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph;

[0177] Assigning a time coordinate to each target prediction map based on the treatment information, where the time coordinate is the expected time corresponding to when the affected part recovers to the stage matching the target prediction map;

[0178] The target prediction map is associated with the time axis according to the corresponding time coordinate to generate a rehabilitation prediction atlas of the affected part.

[0179] In some embodiments, the output of the objective function is proportional to the expected value of the segmentation score and inversely proportional to the update step size of the second segmentation model. When the processor 301 trains the second segmentation model according to the sentence dataset until the output of the objective function converges, the processor 301 includes:

[0180] The second segmentation model is copied to generate a third segmentation model, and the update step size is the strategy difference between the segmentation points of the model generated by the second segmentation model and the segmentation points of the model generated by the third segmentation model;

[0181] Optimize the strategy of generating model segmentation points of the third segmentation model based on the sentence dataset;

[0182] The model parameters of the third segmentation model are copied to the second segmentation model, so that the second segmentation model adopts the same strategy of generating model segmentation points as the third segmentation model until the output of the objective function converges.

[0183] In some embodiments, the correction constraint includes at least a recovery degree constraint and a recovery speed constraint of the affected part;

[0184] When the processor 301 performs a prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph, the processor 301 includes:

[0185] Extracting lesion features from the state prediction graph;

[0186] The lesion features are modified according to the recovery degree constraint and the recovery speed constraint to obtain the modified features;

[0187] Reconstruct and generate a state correction map based on the correction feature;

[0188] The state correction graph is filtered according to the recovery degree constraint to obtain the target prediction graph.

[0189] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. When the processor 301 generates the modified constraints corresponding to the state prediction graph based on the treatment information, the processor 301 includes:

[0190] Generate recovery level constraints based on health status;

[0191] Generate recovery speed constraints based on health status and treatment plan.

[0192] In some embodiments, the lesion feature includes a location feature and a first contour feature:

[0193] When the processor 301 corrects the lesion feature according to the recovery degree constraint and the recovery speed constraint to obtain the corrected feature, the processor 301 includes:

[0194] Generate a regression coefficient based on the recovery speed constraint and the position characteristics;

[0195] Converting the first contour feature into a first area feature value and a first morphological feature value;

[0196] A second contour feature is generated according to the fading coefficient, the first area feature value, and the first morphological feature value to obtain a modified feature including a position feature and a second contour feature.

[0197] In some embodiments, the treatment information includes at least the patient's health status and the treatment plan adopted by the patient. When the processor 301 assigns a time coordinate to each target prediction graph based on the treatment information, the processor 301 includes:

[0198] Determine the target degree of recovery based on the health status and the target total period required to achieve the target degree of recovery under the conditions of the health status and treatment plan;

[0199] Determine the sub-cycles required to achieve the degree of recovery corresponding to the target prediction graph based on the treatment plan;

[0200] The time coordinate is assigned to the corresponding target prediction map according to the relative proportion of the sub-cycle to the target total cycle.

[0201] In some embodiments, when the processor 301 analyzes the scan data to extract the affected area features, the processor 301 includes:

[0202] generating first diagnosis and treatment information according to the scan data;

[0203] Obtaining the diagnosis and treatment evaluation input by the physician, and performing semantic recognition on the diagnosis and treatment evaluation to generate second diagnosis and treatment information;

[0204] An affected area feature is generated according to the first diagnosis and treatment feature and the second diagnosis and treatment feature.

[0205] In some embodiments, when the processor 301 reconstructs the human affected area model according to the affected area characteristics, the processor 301 includes:

[0206] Determine the corresponding affected area according to the characteristics of the affected area;

[0207] An affected part reference model corresponding to the affected part is called, and parameters of the affected part reference model are adjusted according to the affected part characteristics to generate a human affected part model.

[0208] It should be noted that those skilled in the art can clearly understand that, for the sake of convenience and brevity of description, the specific working process of the computer device described above can refer to the corresponding process in the aforementioned embodiment of the affected part rehabilitation prediction method, and will not be repeated here.

[0209] An embodiment of the present application also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any of the affected area rehabilitation prediction methods provided in the embodiments of the present application specification.

[0210] The storage medium may be an internal storage unit of the computer device in the aforementioned embodiment, such as a hard disk or memory of the computer device. The storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device.

[0211] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As is well known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable, and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media generally embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0212] It should be understood that the term "and / or" used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system that includes a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements that are inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system that includes the element.

[0213] The serial numbers of the embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments. The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for predicting the recovery of an affected part, characterized in that: The method comprises: acquiring scan data obtained by scanning an affected part of a patient, and analyzing the scan data to extract affected part features of the affected part; reconstructing a human affected area model according to the affected area characteristics; Generate a prediction diagram of the state of the affected part in each stage of the rehabilitation process based on the human affected part model; Acquiring treatment information of the patient, and generating a correction constraint corresponding to the state prediction graph based on the treatment information; Performing prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph; Assigning a time coordinate to each of the target prediction graphs according to the treatment information, wherein the time coordinate is the expected time corresponding to when the affected part recovers to a stage matching the target prediction graph; Associating the target prediction graph with a time axis according to corresponding time coordinates to generate a rehabilitation prediction atlas of the affected part; The treatment information includes at least the health status of the patient and the treatment plan adopted by the patient, and the correction constraint includes the recovery degree constraint and the recovery speed constraint of the affected part; Generating a correction constraint corresponding to the state prediction graph according to the treatment information includes: generating the recovery degree constraint according to the health condition; generating the recovery speed constraint according to the health condition and the treatment plan; Assigning a time coordinate to each target prediction map according to the treatment information includes: determining a target degree of recovery based on the health condition, and determining a target total period required to achieve the target degree of recovery under the conditions of the health condition and the treatment plan; determining, based on the treatment plan, a sub-cycle required to achieve the degree of recovery corresponding to the target prediction graph; The time coordinate is assigned to the corresponding target prediction graph according to the relative ratio of the sub-cycle to the target total cycle.

2. The method for predicting the recovery of an affected part according to claim 1, wherein: The performing prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph includes: Extracting lesion features in the state prediction graph; Correcting the lesion feature according to the rehabilitation degree constraint and the rehabilitation speed constraint to obtain a corrected feature; Reconstruct and generate a state correction map based on the correction feature; The state correction graph is filtered according to the recovery degree constraint to obtain the target prediction graph.

3. The method for predicting the recovery of an affected part according to claim 2, wherein: The lesion features include position features and first contour features: Correcting the lesion feature according to the rehabilitation degree constraint and the rehabilitation speed constraint to obtain a corrected feature includes: generating a fading coefficient according to the rehabilitation speed constraint and the position feature; Converting the first contour feature into a first area feature value and a first morphological feature value; A second contour feature is generated according to the fading coefficient, the first area feature value, and the first morphological feature value to obtain the modified feature including the position feature and the second contour feature.

4. The method for predicting the recovery of an affected part according to any one of claims 1 to 3, characterized in that: The analyzing the scan data to extract the affected area features of the affected area includes: generating first diagnosis and treatment information according to the scan data; Obtaining a diagnosis and treatment evaluation input by a physician, and performing semantic recognition on the diagnosis and treatment evaluation to generate second diagnosis and treatment information; The affected area feature is generated according to the first diagnosis and treatment information and the second diagnosis and treatment information.

5. The method for predicting the recovery of an affected part according to any one of claims 1 to 3, characterized in that: The reconstructing of the human affected part model according to the affected part characteristics includes: determining a corresponding affected area according to the affected area characteristics; An affected part reference model corresponding to the affected part is called, and parameters of the affected part reference model are adjusted according to the affected part characteristics to generate the human body affected part model.

6. A device for predicting the recovery of an affected part, characterized in that: include: a feature extraction module, configured to obtain scan data obtained by scanning an affected part of a patient, and analyze the scan data to extract affected part features of the affected part; A model reconstruction module, used for reconstructing a human affected part model according to the affected part characteristics; A state prediction module, configured to generate a state prediction diagram of the affected part at each stage of the rehabilitation process based on the human affected part model; a constraint generation module, configured to obtain treatment information of the patient and generate a correction constraint corresponding to the state prediction graph based on the treatment information; A prediction correction module, configured to perform prediction correction on the state prediction graph according to the correction constraint to generate a corresponding target prediction graph; a time prediction module, configured to assign a time coordinate to each of the target prediction graphs based on treatment information, wherein the time coordinate is the expected time corresponding to when the affected part recovers to a stage matching the target prediction graph; An atlas generation module, configured to associate the target prediction map with a time axis according to corresponding time coordinates to generate a rehabilitation prediction atlas of the affected part; The treatment information at least includes the health status of the patient and the treatment plan adopted by the patient, and the correction constraint includes the recovery degree constraint and the recovery speed constraint of the affected part; Generating a correction constraint corresponding to the state prediction graph according to the treatment information includes: generating the recovery degree constraint according to the health condition; generating the recovery speed constraint according to the health condition and the treatment plan; Assigning a time coordinate to each target prediction map according to the treatment information includes: determining a target degree of recovery based on the health condition, and determining a target total period required to achieve the target degree of recovery under the conditions of the health condition and the treatment plan; determining, based on the treatment plan, a sub-cycle required to achieve the degree of recovery corresponding to the target prediction graph; The time coordinate is assigned to the corresponding target prediction graph according to the relative ratio of the sub-cycle to the target total cycle.

7. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and implement the affected part rehabilitation prediction method according to any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the affected part rehabilitation prediction method according to any one of claims 1 to 5.

Images