Rehabilitation plan adjustment method, device, equipment and medium based on user portrait
Through the rehabilitation plan adjustment method based on user portraits, and the preset model is used to dynamically adjust the rehabilitation plan, the problem of lack of personalized and dynamic adjustment of rehabilitation plans in the existing technology is solved, and the rehabilitation effect is improved.
Patent Information
- Application Number
- CN202410639196.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-22
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-05-22
AI Technical Summary
The existing orthopedic rehabilitation plan lacks personalized and dynamic adjustments, and cannot effectively adapt to changes in patients' rehabilitation process, resulting in unsatisfactory rehabilitation results.
The rehabilitation plan adjustment method based on user portraits, by obtaining actual user portrait data and feedback data, using the preset rehabilitation plan adjustment model to dynamically adjust the rehabilitation plan, considering the degree and results of the patient's rehabilitation measures at different stages, and generating personalized and targeted the final recovery plan.
The personalized and dynamic adjustment of the rehabilitation plan has been achieved, the rehabilitation effect has been improved, the changes in patients' rehabilitation process have been adapted, and the pertinence and adaptability of the rehabilitation plan has been improved.
Smart Images

Figure CN118335283B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical health technology, and in particular to a method, device, equipment and medium for adjusting a rehabilitation plan based on user portraits. Background Art
[0002] During orthopedic rehabilitation, a rehabilitation plan is usually developed for the patient based on the patient's admission, treatment, and other data. The patient undergoes rehabilitation training in accordance with the pre-established rehabilitation plan while in the hospital or after discharge. The hospital will regularly revisit the patient to determine the patient's recovery status.
[0003] However, the above-mentioned rehabilitation plans are formulated based on fixed initial admission and treatment conditions. Patients often face problems such as insufficient personalization, lack of specificity, and lack of dynamic adjustment in rehabilitation plans. Existing rehabilitation plans are mostly based on fixed templates, which are difficult to adapt to changes in the patient's rehabilitation process, resulting in unsatisfactory rehabilitation results. Summary of the Invention
[0004] In view of this, the present invention provides a rehabilitation plan adjustment method, device, equipment and medium based on user portraits to solve the problem that rehabilitation plans in the prior art are formulated based on fixed initial admission conditions and treatment conditions, and patients often face the problem that rehabilitation plans are not personalized enough, lack specificity and dynamic adjustment.
[0005] In a first aspect, the present invention provides a rehabilitation plan adjustment method based on user portraits, and the rehabilitation plan adjustment method based on user portraits includes: obtaining measured user portrait data of a target user; determining the target user's initial rehabilitation plan and target follow-up content based on the measured user portrait data; obtaining measured follow-up feedback data of the target user in response to feedback on the target follow-up content; determining the target user's actual rehabilitation status based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of actual execution degrees of phased rehabilitation measures and a corresponding sequence of actual phased rehabilitation results; when the actual rehabilitation status does not reach a warning state, the sequence of actual execution degrees of phased rehabilitation measures and / or the sequence of actual phased rehabilitation results are input into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user.
[0006] As an exemplary embodiment, when the actual rehabilitation status has not reached the warning status, the actual execution degree sequence of the staged rehabilitation measures and / or the actual rehabilitation result sequence of the staged rehabilitation measures are input into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user, including: extracting the temporal dependency relationship of the actual execution degree sequence of the staged rehabilitation measures to obtain a first temporal change feature; the first temporal change feature is used to represent the dynamic characteristics of the actual execution degree of the rehabilitation measures between different stages of the rehabilitation plan; extracting the temporal dependency relationship of the staged actual rehabilitation result sequence to obtain a second temporal change feature; the second temporal change feature is used to represent the dynamic characteristics of the actual rehabilitation results between different stages of the rehabilitation plan; fusing the first temporal change feature and the second temporal change feature to obtain a temporal accumulation feature; wherein the temporal accumulation feature is used to represent the actual rehabilitation status of the target user; inputting the first temporal change feature, the second temporal change feature and the temporal accumulation feature into the preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user.
[0007] As an exemplary embodiment, the measured follow-up feedback data is subjected to temporal dependency extraction to obtain a first temporal change feature and a second temporal change feature, including: determining the target user's part to be rehabilitated based on the measured user portrait; extracting the measured follow-up feedback data based on the part to be rehabilitated to obtain a rehabilitation part feature of the part to be rehabilitated; and extracting the temporal dependency of the rehabilitation part feature to obtain a first temporal change feature and a second temporal change feature.
[0008] As an exemplary embodiment, the time series dependency of the measured follow-up feedback data is extracted to obtain a first time series change feature and a second time series change feature, and the method further includes: extracting a training activity feature that characterizes the user's rehabilitation training activity status from each measured follow-up feedback data; determining the actual rehabilitation training activity status of the target user based on the training activity feature; wherein the actual rehabilitation training activity status includes at least one of exercise frequency, intensity, and duration; and extracting a time series dependency of the actual rehabilitation training activity status to obtain a first time series change feature and a second time series change feature.
[0009] As an exemplary embodiment, the actual execution degree sequence of staged rehabilitation measures and / or the actual rehabilitation result sequence of stages are input into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user, including: extracting the rehabilitation static feature sequence of the actual execution degree sequence of staged rehabilitation measures and / or the actual rehabilitation result sequence of stages; determining the preset rehabilitation feature sequence that is time-aligned with the rehabilitation static feature sequence based on the preset rehabilitation time node sequence in which the rehabilitation static feature sequence is located; constructing a rehabilitation difference feature sequence based on the rehabilitation static feature sequence and the preset rehabilitation feature sequence; and inputting the rehabilitation difference feature sequence into the pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user.
[0010] As an exemplary embodiment, the method also includes: when the actual rehabilitation status reaches the warning state, the measured user portrait data, the measured follow-up feedback data, and the initial rehabilitation plan are input into a pre-trained warning prediction model to obtain warning prompt information, and the warning prompt information is used to warn the target user.
[0011] As an exemplary embodiment, a method for constructing a preset rehabilitation plan adjustment model includes: obtaining multiple groups of historical user portrait data, historical initial rehabilitation plan data, historical follow-up feedback data, historical rehabilitation plan adjustment data and historical rehabilitation status of the target user as an initial data set; as an exemplary embodiment, dividing the initial data set based on the historical rehabilitation status to obtain a first historical rehabilitation data set; wherein the historical rehabilitation status in the first historical rehabilitation data set conforms to the first preset rehabilitation status; extracting the temporal dependency relationship of the historical follow-up feedback data to obtain a historical first temporal change feature and a historical second temporal change feature; wherein the historical first temporal change feature is used to represent the historical execution degree of the rehabilitation measures, and the historical second temporal change feature is used to represent the stage-by-stage historical results of the rehabilitation plan of the target user; fusing the historical first temporal change feature and the historical second temporal change feature to obtain a historical temporal cumulative feature; wherein the temporal cumulative feature is used to represent the actual rehabilitation status of the target user; inputting the historical first temporal change feature, the historical second temporal change feature, the historical temporal cumulative feature and the corresponding historical rehabilitation plan adjustment data into the initial rehabilitation plan adjustment model, and performing model training with the historical rehabilitation plan adjustment data as the target value to obtain a preset rehabilitation plan adjustment model.
[0012] In a second aspect, the present invention provides a rehabilitation plan adjustment device based on user portraits, and the rehabilitation plan adjustment device based on user portraits includes: a first acquisition module for acquiring measured user portrait data of a target user; a determination module for determining the initial rehabilitation plan and target follow-up content of the target user based on the measured user portrait data; a second acquisition module for acquiring measured follow-up feedback data of the target user in response to feedback on the target follow-up content; a rehabilitation status determination module for determining the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of actual execution degrees of staged rehabilitation measures and a corresponding sequence of actual staged rehabilitation results; a rehabilitation plan generation module for inputting the sequence of actual execution degrees of staged rehabilitation measures and / or the sequence of actual staged rehabilitation results into a pre-trained preset rehabilitation plan adjustment model when the actual rehabilitation status does not reach the warning state, so as to obtain the final rehabilitation plan for the target user.
[0013] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to execute the user portrait-based rehabilitation plan adjustment method of the above-mentioned first aspect or any corresponding embodiment thereof.
[0014] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the user portrait-based rehabilitation plan adjustment method of the above-mentioned first aspect or any corresponding embodiment thereof.
[0015] The above implementation method of this embodiment obtains the measured user portrait data of the target user; determines the initial rehabilitation plan and target follow-up content of the target user based on the measured user portrait data; obtains the measured follow-up feedback data of the target user in response to the target follow-up content feedback; determines the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of the actual execution degree of staged rehabilitation measures and a corresponding sequence of actual rehabilitation results of the stage; when the actual rehabilitation status does not reach the warning state, the sequence of the actual execution degree of the staged rehabilitation measures and / or the sequence of actual rehabilitation results of the stage are input into a pre-trained preset rehabilitation plan adjustment model The above implementation method, firstly, can customize the initial rehabilitation plan for the target user through the measured user portrait data; secondly, the implementation degree of the rehabilitation measures contained in the measured follow-up feedback data and the actual rehabilitation results at each stage can be used to determine the implementation status of the target user during rehabilitation training based on the initial rehabilitation plan, so as to take into account the differences in subjective and objective factors of the target user when executing the rehabilitation plan. In each rehabilitation stage, the final rehabilitation plan for the implementation status can be continuously and adaptively generated through the preset rehabilitation plan adjustment model, so as to dynamically adjust the rehabilitation plan in a personalized and targeted manner to adapt to the rehabilitation status of the target user. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0017] Figure 1 is a flowchart of a method for adjusting a rehabilitation plan based on a user portrait according to an embodiment of the present invention;
[0018] Figure 2 is a flowchart of another method for adjusting a rehabilitation plan based on a user portrait according to an embodiment of the present invention;
[0019] Figure 3 is a structural block diagram of a rehabilitation plan adjustment device based on user portrait according to an embodiment of the present invention;
[0020] Figure 4 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0021] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.
[0022] In a first aspect, an embodiment of the present invention provides a method for adjusting a rehabilitation plan based on a user profile. Figure 1 FIG. 1 is a flow chart of a method for adjusting a rehabilitation plan based on user portraits according to an embodiment of the present invention. Figure 1 As shown in the figure, the rehabilitation plan adjustment method based on user portrait includes:
[0023] Step S101: Acquire measured user portrait data of a target user; wherein the measured user portrait data includes at least one of hospital admission cases and treatment data;
[0024] In this embodiment, the measured user portrait data of the target user is constructed based on the information data of the target user collected in advance; wherein, the information data may include dimensions such as basic information, medical history, fracture details, treatment methods, rehabilitation stage, psychological state, social support and rehabilitation methods, which can be determined through the electronic medical records entered by the user when admitted to the hospital.
[0025] Specifically, since patients of different age groups may face different challenges during the rehabilitation process, men and women may differ in physiological structure and rehabilitation needs, occupational activity level and work environment have an impact on rehabilitation plans and expected recovery time, and different dietary habits such as eating habits, exercise habits, whether or not there is fitness or sports training experience, smoking and drinking habits will also lead to differences in the implementation of rehabilitation plans. Based on this, the basic information includes at least: age information, gender information, occupation information and lifestyle information.
[0026] In addition, different rehabilitation plans should be set up for fracture patients with different medical histories. For example, whether the patient has had fractures in similar or different locations will result in differences in rehabilitation. The patient's chronic diseases and long-term use of certain specific medications will also have an impact on rehabilitation. Based on this, the medical history should at least include previous fracture history, chronic diseases, drug use history, etc.
[0027] In addition, different rehabilitation plans need to be set up for fracture patients with different fracture details; for example, different fracture sites, different fracture types and different fracture causes require different rehabilitation methods, and the fracture details can usually be clearly reflected in auxiliary examinations; based on this, the fracture details at least include the fracture site, fracture type, fracture cause, and auxiliary examinations; among them, fracture types include simple fractures, open fractures, and compression fractures, fracture causes include falls, traffic accidents, sports injuries, etc., and auxiliary examinations include X-ray examinations, magnetic resonance imaging examinations, and computed tomography.
[0028] In addition, different rehabilitation plans need to be set up for different treatment methods; for example, different types of surgery, different non-surgical treatment methods and different physical therapy methods will all have an impact on the rehabilitation plan; based on this, the treatment method at least includes surgical type, non-surgical treatment, physical therapy, etc.; among them, surgical types include internal fixation, external fixation, joint replacement, etc.; non-surgical treatments include plaster fixation, traction, etc.; physical therapy includes electrical stimulation, heat therapy, cold therapy, etc.
[0029] In addition, different rehabilitation plans should be set up for different rehabilitation stages; specifically, the rehabilitation stages include the acute phase, the recovery phase, and the functional recovery phase; among them, the acute phase is the initial treatment and pain management after the fracture; the recovery phase aims to gradually increase the range of motion and weight-bearing as the fracture heals; the functional recovery focuses on restoring the patient's normal function and quality of life.
[0030] In addition, different rehabilitation plans should be set up for different social supports; specifically, the social support includes family environment and social resources; among them, the family environment refers to the support of family members and the adaptability of the family environment, and social resources refer to available resources such as rehabilitation centers and community services.
[0031] Finally, different rehabilitation plans should be set up according to different rehabilitation preferences; specifically, the rehabilitation preferences include rehabilitation goals and rehabilitation methods; among them, rehabilitation goals are the patient's expectations for rehabilitation outcomes, such as complete recovery, partial recovery, etc.; rehabilitation methods are the patient's preferences for rehabilitation methods, such as home rehabilitation, outpatient rehabilitation, etc.
[0032] Step S102: Determine the target user's initial rehabilitation plan and target follow-up content based on the measured user portrait data.
[0033] In this embodiment, the initial rehabilitation plan may be in the form of a rehabilitation course; wherein, the rehabilitation course may include multiple preset rehabilitation nodes, and in each rehabilitation node, the target person needs to perform phased rehabilitation measures within the preset rehabilitation time according to the preset rehabilitation frequency and preset rehabilitation intensity.
[0034] In this embodiment, the initial rehabilitation plan and target follow-up content of the target user are first determined through the measured user portrait data containing multi-dimensional data; specifically, the initial rehabilitation plan and follow-up content database can be set in advance, and the initial rehabilitation plan and target follow-up content of the target user corresponding to the measured user portrait data are determined through the similarity between the measured user portrait data and the historical user portrait data in the initial rehabilitation plan and follow-up content database; when performing the similarity comparison, focus is placed on the data in the user portrait data, such as fracture details, treatment methods, rehabilitation stages, and other dimensions that are highly correlated with the disease condition.
[0035] Among them, the initial rehabilitation plan and follow-up content database can be constructed based on the historical user portrait data, historical initial rehabilitation plan and historical follow-up content generated in the historical diagnosis and treatment process collected in advance; specifically, the historical user portrait data, historical initial rehabilitation plan and historical follow-up content generated in the historical diagnosis and treatment process are collected in advance, and after collecting the historical user portrait data, historical initial rehabilitation plan and historical follow-up content, a corresponding database is established based on the correspondence between the historical user portrait data and the historical initial rehabilitation plan and the correspondence between the historical user portrait data and the historical follow-up content.
[0036] In this embodiment, the target follow-up content can be pushed to the target user via a questionnaire; the questionnaire at least includes time information, training status information, personal feeling information, etc. to be uploaded by the user; among them, the user can upload the training status information in the form of text, image, or video.
[0037] For example, for a target user undergoing femoral neck fracture and hip replacement, an initial rehabilitation plan with an eight-week recovery period can be set. Within this initial rehabilitation plan, different phased rehabilitation measures can be set. Specifically, the initial rehabilitation plan and target follow-up content can be:
[0038] Rehabilitation nodes during the acute treatment period from postoperative to the fifth day after surgery:
[0039] Phased rehabilitation measures: 1. Ankle pumps: Lie on your back with your heels elevated 20cm. Apply slight force to perform ankle dorsiflexion, hold for 10 seconds, then slowly perform ankle flexion for 10 seconds. Do 20 reps every hour. 2. Quadriceps isometric contraction: Lie on your back, tighten the muscles in front of your thighs, press your knees down against the bed, hold for 10 seconds, and slowly relax. Do 12 reps in a group, 3 sets at a time, rest for 30 seconds between each set, twice a day. 3. Gluteal isometric contraction: Lie on your back, squeeze your buttocks (the pelvis will be slightly raised when the muscles on both sides are contracted), hold for 5 seconds, do 12 reps in a group, 3 sets at a time, rest for 30 seconds between each set. Twice a day; 4. Lower limb mobility training: In the supine position, slide your heels along the bed toward your buttocks until your calves are flexed to 45 degrees. Hold for 5 seconds, then slowly straighten your calves along the bed. Do 10 reps in a group, 3 sets each time, twice a day; 5. Hip abduction training: In the supine position, slowly extend your lower limbs horizontally to the sides along the bed, then slowly retract them. Do 10 reps in a group, 2-3 sets each time, with a 30-second rest between each set. 2-3 times a day; 6. Quadriceps training: In a seated position (lean back slightly, with the angle between your thighs and abdomen no more than 90 degrees), perform a full range of motion for your calves, holding for 5 seconds at the end of the extension. Do 12 reps in a group, 3 sets each time, with a 30-second rest between each set. 2 times a day (you can tie a sandbag according to your tolerance); 7. Standing leg lift: In a standing position (with support in front), lift your thighs alternately with both lower limbs (no more than 90°), hold for 3 seconds at the end of the lift, and then slowly lower them. Do 12 in a group, 3 groups each time, and rest for 30 seconds between each group. 2 times a day; 8. Standing hip extension training: In a standing position, extend the hip outward and backward no more than 15°, hold for 3 seconds, and then slowly retract it. Do 12 in a group, 3 groups each time, and rest for 30 seconds between each group. 2 times a day; 9. Standing hip abduction training: In a standing position, abduct the hip, hold for 3 seconds, and then slowly retract it. Do 12 in a group, 3 groups each time, and rest for 30 seconds between each group. Twice a day; 10. Micro-squat training: Stand with your feet shoulder-width apart, toes and knees facing forward, and slowly bend your knees 30 degrees (toes should not exceed knees, back straight). Hold for 10 seconds, then slowly stand up. Do 12 sets, 3 sets at a time, resting 30 seconds between each set. Twice a day.
[0040] The target follow-up content for this phased rehabilitation measure can be: prompting the target user to upload the execution status of each action within the acute treatment period rehabilitation node; prompting the target user to upload the target patient's feelings of the affected part, as well as image information and video information at the acute treatment period rehabilitation node; prompting the target user to upload the actual rehabilitation status within the acute treatment period rehabilitation node, for example, whether the edema is reduced, the pain in the affected part, and whether independent transfer and safe getting in and out of bed, chair, and toilet can be achieved, and independent walking on flat ground with a walker can be achieved; pushing precautions for total hip replacement to the target user; among which, the precautions can be: avoid hip flexion exceeding 90°, adduction exceeding the midline, internal rotation exceeding the neutral position, avoid sitting for a long time for no more than 1 hour at a time, avoid placing a pillow under the knees to prevent hip flexion contracture, and use an abduction pillow when lying in the supine position.
[0041] The second postoperative phase, from 2 to 10 weeks after surgery:
[0042] Phased rehabilitation measures: 1. Shell-like exercise: Lie on your side, bend your knees, with the affected side on top, slowly open your knees upward (like an open shell), hold at the end for 5 seconds, then slowly lower them, 12 in a group, do 3 groups each time, rest for 30 seconds between each group, twice a day; Standing hip abduction training: In a standing position, slowly extend one thigh outward (keep the buttocks still, tighten the abdomen, and do not twist the body), hold at the end for 5 seconds, then slowly retract them, 12 in a group, do 2 groups each time, rest for 30 seconds between each group. 2. Standing hip extension training: In a standing position, extend the hip joint outward and backward no more than 15°, hold for 3 seconds, and then slowly retract it. Do 12 in a group, do 3 groups each time, rest for 30 seconds between each group, twice a day; Bridge exercise: 3. In a supine position, bend the knees of both lower limbs 90°, slowly lift the buttocks off the bed, hold the end for 5 seconds, and then slowly retract it. Do 12 in a group, do 5 groups each time, rest for 30 seconds between each group, 2-3 times a day; 4. Micro squat training: In a standing position, with your feet shoulder-width apart, toes and knees facing forward, slowly bend your knees 30° (toes should not exceed knees, back straight), hold for 10 seconds, and then slowly stand up. Do 12 in a group, do 3 groups each time, rest for 30 seconds between each group, twice a day.
[0043] The target follow-up content for this stage of rehabilitation measures can be: prompting the target user to upload the execution status of each action in the second stage; prompting the target user to upload the feeling of the affected area in the second stage rehabilitation node as well as image information and video information; prompting the target user to upload the actual rehabilitation status in the acute treatment rehabilitation node, for example, whether the gait is normalized with the assistive device, and whether daily activities can be performed independently; pushing the precautions for the second stage to the target user; among which the precautions can be: avoid hip flexion exceeding 90°, adduction exceeding the midline, internal rotation exceeding the neutral position, sitting for no more than 1 hour at a time, using an abduction pillow when lying on the back, and avoiding climbing stairs with alternating legs.
[0044] The third phase is from 10 to 14 weeks after surgery:
[0045] Phased rehabilitation measures: 1. Lateral step training: In a standing position, stand on the step with the affected leg, flex the hip 20°, slowly lower the healthy leg from the step, and then slowly lift it up (keep your head, chest and abdomen lifted), 12 reps in a group, 3 sets each time, 30 seconds rest between each set. Twice a day; 2. Alternating up and down step exercises: Alternate up and down the steps with both feet (keep your head, chest and abdomen lifted), 12 reps in a group, 2 sets each time, 30 seconds rest between each set. Twice a day; 3. Wall rolling ball micro squat training: In a standing position, lean against the wall slightly (feet shoulder-width apart), clamp a ball between your back and the wall, slowly squat, slowly bend your knees 30° (toes do not exceed knees, back straight), hold for 10 seconds, then slowly stand up, 12 reps in a group, 3 sets each time, 30 seconds rest between each set. Twice a day; 4. Bridge exercise: supine position, flex both lower limbs at 90 degrees, slowly lift the buttocks off the bed, hold the end for 5 seconds, then slowly retract, 12 sets, 3 sets each time, rest 30 seconds between each set. Twice a day; 5. Straight leg raise: supine position, flex the lower limb on the healthy side at 90 degrees, with the lower limb on the affected side extended, lift the thigh 20 to 30 cm off the bed, hold the end for 5 seconds, then slowly retract, 12 sets, 3 sets each time, rest 30 seconds between each set. Twice a day; 6. Balance / proprioception training: dynamic activities of both legs and one leg (15 to 20 minutes); 7. Power bike (20 to 30 minutes, you can go to the gym); 8. Bilateral leg kick / unilateral leg kick exercise (15 sets, 3 to 5 sets each time, you can go to the gym).
[0046] The target follow-up content for this stage of rehabilitation measures can be: prompting the target user to upload the execution status of each action in the third stage; prompting the target user to upload the feelings of the affected area in the third stage as well as image information and video information; prompting the target user to upload the actual rehabilitation status in the rehabilitation node of the acute treatment period, for example, whether the target user can complete alternating up and down stairs, whether the target user can independently complete the lower body dressing, including putting on and taking off shoes and socks, and whether the target user can perform special functional activities (cycling, swimming).
[0047] It should be understood that the phased rehabilitation measures and target follow-up contents recorded in the above-mentioned initial rehabilitation plan are only for illustrative purposes of the technical solution in this embodiment, and do not specifically limit the scope of protection of this embodiment; among them, the above-mentioned phased rehabilitation measures can be adaptively adjusted in terms of increase or decrease and time of rehabilitation measures based on the measured user portrait data, and the above-mentioned target follow-up contents can be adaptively generated based on the adaptively adjusted phased rehabilitation measures.
[0048] Step S103: obtaining the measured follow-up feedback data of the target user in response to the target follow-up content feedback.
[0049] In this embodiment, the measured follow-up feedback data of the target in response to the target follow-up content feedback may include images, text, videos, etc.; among them, the text can be the text information filled out by the user in response to the questionnaire, and the text information is used to indicate the implementation status or rehabilitation results of the rehabilitation measures; the images and videos can be the image data and video data of the part to be rehabilitated taken by the user in response to the questionnaire.
[0050] Step S104, determining the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series sequence of actual implementation levels of phased rehabilitation measures and a corresponding sequence of actual rehabilitation results at each phase.
[0051] Since each target user has different subjective and objective factors during the rehabilitation process, such as psychological state, willingness to perform rehabilitation movements, family environment and social resources, even if an initial rehabilitation plan is customized for the target user based on the measured follow-up feedback data, there will be deviations during execution, and this deviation will lead to different actual rehabilitation states; therefore, in this embodiment, the actual rehabilitation state of the target user is determined based on the initial rehabilitation plan and the measured follow-up feedback data; as a possible implementation method, historical rehabilitation data with known rehabilitation results can be collected in advance, and the actual rehabilitation state can be determined according to the similarity between the initial rehabilitation plan and the measured follow-up feedback data and the historical rehabilitation data.
[0052] Among them, the historical rehabilitation data may include multiple groups of historical user portrait data, historical initial rehabilitation plan data, historical follow-up data and historical rehabilitation result data; specifically, after collecting the historical rehabilitation result data, the historical rehabilitation data is first classified based on the historical rehabilitation result data to obtain a first historical rehabilitation data set and a second historical rehabilitation data set; wherein, the first historical rehabilitation data set is data whose historical rehabilitation results have not reached a warning state; the second historical rehabilitation data set is data whose historical rehabilitation results have reached a warning state; after dividing the second historical rehabilitation data set, the historical rehabilitation status corresponding to the historical user portrait data and the historical follow-up feedback data with a similarity greater than a preset degree can be used as the measured rehabilitation status of the target user through the degree of similarity between the measured user portrait data and its corresponding measured follow-up feedback data and the historical user portrait data contained in the first historical rehabilitation data set and the historical user portrait data contained in the second historical rehabilitation data set and its corresponding historical follow-up feedback data.
[0053] When the target user undergoes rehabilitation training according to the initial rehabilitation plan generated based on the measured user portrait data, due to objective factors such as his basic information and physical condition, or subjective factors such as his psychological state and social support, the implementation of the phased rehabilitation measures during rehabilitation training will be different from expectations, which will lead to deviations in the actual rehabilitation status of the target user. When the user's actual rehabilitation status deviates, it usually leads to two results. One is that although there is a deviation, adjusting the subsequent rehabilitation plan based on the deviation can achieve a better prognosis. This usually occurs when the target user has a high degree of cooperation and can better execute the first preset rehabilitation level of the corresponding rehabilitation plan; or, the target user has a high degree of cooperation, but although he can execute the rehabilitation plan, there are certain differences in the execution from the initial rehabilitation plan. Deviation is the actual rehabilitation state that has not reached the warning rehabilitation state, but the deviation can be controlled within an acceptable range; the other is the actual rehabilitation state that has reached the warning rehabilitation state with a large deviation, even if the subsequent rehabilitation plan is adjusted according to the deviation, the expected rehabilitation effect cannot be achieved, and may even lead to new or irreversible injuries or even more serious injuries and the occurrence of other results; for the target user who has not reached the warning rehabilitation state, the initial rehabilitation plan can be adaptively adjusted according to the actual measured user portrait and the actual measured follow-up feedback data; and for the target user of the second recovery level, the possible results of continuing to implement the initial rehabilitation plan according to the existing status can be adaptively predicted according to the actual measured user portrait and the actual measured follow-up feedback data, so as to provide warnings to such target users as much as possible.
[0054] Step S105: When the actual rehabilitation status does not reach the warning state, the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence are input into the pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user.
[0055] When the actual rehabilitation status does not reach the warning status, the initial rehabilitation plan needs to be adjusted; in this embodiment, the above adjustment process is implemented through a preset rehabilitation plan adjustment model.
[0056] As a possible implementation method, the rehabilitation plan execution process and rehabilitation results of the target user can be regarded as a form of time series accumulation; specifically, in each rehabilitation stage, the accumulation of the rehabilitation measures before the rehabilitation stage and the execution degree of the rehabilitation measures in the rehabilitation stage over time determines the actual rehabilitation result of the stage in the rehabilitation stage, and the accumulation of all the stage rehabilitation measures currently performed by the target user over time, as well as the accumulation of the actual rehabilitation results of the stage, determine the final rehabilitation result of the target user; based on this, in this embodiment, in order to comprehensively consider the temporal changes of the execution degree of the above-mentioned rehabilitation measures and the actual rehabilitation results of the stage, the execution degree and the actual rehabilitation results of the stage based on the rehabilitation measures are extracted from the measured follow-up data, and a time series accumulation feature is constructed; the preset rehabilitation plan adjustment model is used to output the final rehabilitation plan of the target user based on the time series accumulation feature; this process can determine the execution status of the target user during rehabilitation training based on the initial rehabilitation plan through the execution degree of the rehabilitation measures and the actual rehabilitation results of the stage contained in the measured follow-up feedback data, and adaptively generate a final rehabilitation plan for the execution status through the preset rehabilitation plan adjustment model.
[0057] As another possible implementation method, the rehabilitation results fed back by the target user can also be regarded as a form of expression represented by static features; specifically, the features of the target user's area to be rehabilitated at the end of the rehabilitation stage can be used to represent the degree of execution of the rehabilitation measures of the target user from the beginning of rehabilitation in accordance with the initial rehabilitation plan to the current rehabilitation stage; that is, the target user's rehabilitation plan can be adjusted based on the target user's stage-by-stage actual rehabilitation results at the end of the rehabilitation stage; based on this, in this embodiment, in order to comprehensively consider the static features of the execution degree of the above-mentioned rehabilitation measures and the stage-by-stage actual rehabilitation results, the stage-by-stage actual rehabilitation results of the target user in the current rehabilitation stage are extracted from the measured follow-up data, and the preset rehabilitation plan adjustment model is used to output the target user's final rehabilitation plan based on the stage-by-stage actual rehabilitation results; this process can determine the execution status of the target user during rehabilitation training based on the initial rehabilitation plan through the stage-by-stage actual rehabilitation results contained in the measured follow-up feedback data, and adaptively generate a final rehabilitation plan for the execution status through the preset rehabilitation plan adjustment model.
[0058] The above implementation method of this embodiment obtains the measured user portrait data of the target user; wherein, the measured user portrait data includes at least one of the hospitalization cases and treatment data; determines the initial rehabilitation plan and target follow-up content of the target user based on the measured user portrait data; obtains the measured follow-up feedback data of the target user in response to the target follow-up content feedback; determines the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of the actual execution degree of staged rehabilitation measures and a corresponding sequence of staged actual rehabilitation results; when the actual rehabilitation status does not reach the warning state, the sequence of the actual execution degree of staged rehabilitation measures and / or the sequence of staged actual rehabilitation results are input into the warning state. The final rehabilitation plan for the target user is obtained from the pre-trained preset rehabilitation plan adjustment model. The above implementation method, first, can customize the initial rehabilitation plan for the target user through the measured user portrait data; secondly, the execution level of the rehabilitation measures contained in the measured follow-up feedback data and the actual rehabilitation results at each stage can be used to determine the execution status of the target user during rehabilitation training based on the initial rehabilitation plan, so as to take into account the differences in subjective and objective factors of the target user when executing the rehabilitation plan. In each rehabilitation stage, the final rehabilitation plan for the execution status can be continuously and adaptively generated through the pre-set rehabilitation plan adjustment model, so as to dynamically adjust the rehabilitation plan in a personalized and targeted manner to adapt to the rehabilitation status of the target user.
[0059] As an exemplary embodiment, Figure 2 FIG. 1 is a flow chart of another method for adjusting a rehabilitation plan based on a user portrait according to an embodiment of the present invention. Figure 2 As shown, in step S105, when the actual rehabilitation status does not reach the warning state, the sequence of actual implementation levels of the staged rehabilitation measures and / or the sequence of actual staged rehabilitation results are input into the pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user, including:
[0060] Step S1051 , extracting the temporal dependency relationship of the sequence of actual execution levels of staged rehabilitation measures to obtain a first temporal variation feature; the first temporal variation feature is used to represent the dynamic features of the actual execution levels of rehabilitation measures between different stages of the rehabilitation plan.
[0061] As a possible implementation method, the above-mentioned temporal dependency extraction function can be realized through a neural network; specifically, it can be realized through a recurrent neural network (RNN), a long short-term memory network (LSTM) or a gated recurrent unit (GRU), etc.; the measured follow-up feedback data passes through the above-mentioned neural network, and the measured execution degree of the rehabilitation measures is extracted to obtain the measured first temporal change characteristics, and the computational measured results of the rehabilitation plan for the target user are used to obtain the measured second temporal change characteristics.
[0062] In this embodiment, the dynamic characteristic of the actual execution degree of the rehabilitation measure may be the change slope of the rehabilitation execution status, for example, the change slope of the rehabilitation execution status of all rehabilitation nodes that have been completed so far.
[0063] Step S1052 , extracting the temporal dependency relationship of the sequence of actual rehabilitation results at different stages to obtain a second temporal variation feature; the second temporal variation feature is used to represent the dynamic characteristics of the actual rehabilitation results between different stages of the rehabilitation plan.
[0064] In this embodiment, the implementation method of extracting the temporal dependency relationship of the staged actual rehabilitation result sequence can be referred to the content of the above-mentioned step S1051, which will not be repeated here.
[0065] In this embodiment, the dynamic feature of the actual rehabilitation result may be a change slope of the actual rehabilitation result, for example, the change slope of the actual rehabilitation results of all rehabilitation nodes that have been executed so far.
[0066] Step S1053 , fusing the first time series variation feature and the second time series variation feature to obtain a time series cumulative feature; wherein the time series cumulative feature is used to represent the actual recovery status of the target user.
[0067] As a possible implementation method, the measured first temporal change feature and the measured second temporal change feature can be fused through an attention mechanism; wherein, different weights can be assigned to each element in the measured first temporal change feature and the measured second temporal change feature, and the weight value depends on the degree of its influence on the actual rehabilitation status of the target user; for example, a relatively high weight can be assigned to the elements in the measured first temporal change feature used to characterize the time, frequency, intensity, etc. of the rehabilitation action, and a relatively low weight can be assigned to, for example, the psychological state; in the measured second temporal change feature, a relatively high weight is assigned to the elements used to characterize the rehabilitation results that are associated with the elements in the measured first temporal change feature that characterize the time, frequency, intensity, etc. of the rehabilitation action, and the others can be assigned a relatively low weight; after determining the weights, the measured first temporal change feature and the measured second temporal change feature are fused based on the above weights using an attention mechanism to obtain a temporal cumulative feature.
[0068] Step S1054 : inputting the first time-series variation feature, the second time-series variation feature, and the time-series cumulative feature into a preset rehabilitation plan adjustment model to obtain a final rehabilitation plan for the target user.
[0069] Through the above-mentioned implementation of this embodiment, the actual execution degree sequence of the staged rehabilitation measures and / or the staged actual rehabilitation result sequence represented by the first time series change feature, the second time series change feature and the time series accumulation feature are input into the pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan of the target user; this step can take into account the accumulation of the execution degree of the rehabilitation measures of the rehabilitation stage and the rehabilitation measures before the rehabilitation stage in each stage reflected in the measured follow-up feedback data of the target user, the accumulation of all the staged rehabilitation measures currently performed by the target user over time, and the accumulation of the target user's actual rehabilitation results at each stage through the pre-trained rehabilitation plan adjustment model, and adaptively adjust the rehabilitation plan to obtain the final rehabilitation plan of the target user that is adjusted to adapt to the training differences when the target user undergoes rehabilitation training based on the initial rehabilitation plan.
[0070] As an exemplary embodiment, the measured follow-up feedback data is subjected to temporal dependency extraction to obtain a first temporal change feature and a second temporal change feature, including: determining the target user's part to be rehabilitated based on the measured user portrait; extracting the measured follow-up feedback data based on the part to be rehabilitated to obtain a rehabilitation part feature of the part to be rehabilitated; and extracting the temporal dependency of the rehabilitation part feature to obtain a first temporal change feature and a second temporal change feature.
[0071] Since the measured follow-up feedback data is presented in the present invention in a variety of forms such as video, image, text, etc., in order to accurately extract the degree of implementation of rehabilitation measures and the actual rehabilitation results at different stages from the measured follow-up feedback data, the target user's part to be rehabilitated is first determined based on the measured user portrait; specifically, the target user's part to be rehabilitated can be determined through the content of the fracture details recorded in the measured user portrait regarding the part to be rehabilitated.
[0072] Furthermore, for the measured follow-up feedback data of video and image types, the rehabilitation site features of the part to be rehabilitated can be extracted by image recognition in combination with the user portrait of the target user.
[0073] For the measured follow-up feedback data of text type, semantic feature extraction can be used in combination with the user portrait of the target user to obtain multiple semantic features in the measured follow-up feedback data based on text attributes, and the preset text used by the user to describe the rehabilitation part features of the part to be rehabilitated can be screened from the semantic features, and then the rehabilitation part features of the part to be rehabilitated can be obtained based on the preset text, so as to further extract the temporal dependency relationship of the rehabilitation part features to obtain the first temporal change feature and the second temporal change feature.
[0074] In the present invention, corresponding wearable devices such as wearable wristbands and IoT devices such as aerobic exercise bikes can be provided to target users based on the data on the area to be rehabilitated in the user profile of the target person. Based on this, as an exemplary embodiment, time series dependency relationships are extracted from the measured follow-up feedback data to obtain first time series change features and second time series change features, further comprising:
[0075] The training activity features representing the user's rehabilitation training activity status are extracted from each measured follow-up feedback data.
[0076] In this embodiment, the training activity feature that characterizes the user's rehabilitation training activity status can be determined solely through motion data transmitted by a wearable device or an IoT device, or can be determined jointly by a wearable device and an IoT device.
[0077] Determining the actual rehabilitation training activity status of the target user based on the training activity characteristics; wherein the actual rehabilitation training activity status includes at least one of exercise frequency, intensity, and duration;
[0078] After obtaining the training activity features, the training activity features are extracted and analyzed to obtain the actual rehabilitation training activity status of the target user.
[0079] The temporal dependency relationship of the actual rehabilitation training activity state is extracted to obtain the first temporal change feature and the second temporal change feature.
[0080] In some embodiments, the extracted user training activity features are time series features of the user's rehabilitation training activities. When extracting the user training activity features, the user's vital sign data during the training activity can also be extracted at the same time, and a vital sign feature sequence aligned with the time series of the training activity features is extracted from the vital sign data, and static features and vital sign time series change features in the time series vital sign feature sequence are extracted; wherein, the vital sign data may include the target user's heart rate data, electromyography data (for example, electromyography data of muscle groups associated with the rehabilitation site), and other vital sign data, and may also include the target user's facial expression feature data, etc.
[0081] Extract the dependency relationship between the temporal change characteristics of the vital signs and the first temporal change characteristics; wherein, the feature extraction model for extracting the dependency relationship between the temporal change characteristics of the vital signs and the first temporal change characteristics is: using the temporal change characteristics of the vital signs extracted using historical data and the first temporal change characteristics to construct a causal transformation graph, and encoding the causal transformation information between nodes in the causal transformation graph based on the convolutional network of the causal transformation graph. After obtaining the temporal change characteristics of the vital signs of the target user and the first temporal change characteristics, the causal transformation information is extracted as the dependency relationship using the above-mentioned pre-trained convolutional network. Based on the dependency relationship, the causal characteristics of the actual implementation degree of the staged rehabilitation measures are determined. After the causal characteristics of the actual implementation degree of the staged rehabilitation measures of the target user are extracted using the dependency relationship between the temporal change characteristics of the vital signs and the first temporal change characteristics, the target user's own situation can be more individually considered when adjusting the preset rehabilitation plan, making the adjusted rehabilitation plan more conducive to the target user's completion.
[0082] In this embodiment, there is a preset correlation between the physical sign characteristics and the training activity characteristics. When the training activity is performed, it needs to be performed within the reasonable physical sign range of the target user. Therefore, in this embodiment, after extracting a certain training activity feature, it can be compared with the physical sign corresponding to the moment. If the physical sign exceeds the reasonable range, the training activity may not be suitable for the target user and can be used as a target feature for subsequent rehabilitation plan adjustments. If the training activity feature does not meet the standard, and the corresponding physical sign is within the reasonable range, or is lower than the lower limit of the reasonable range; it indicates that the target user's execution willingness is poor. Therefore, when adjusting the subsequent rehabilitation plan, it is necessary to consider improving the target user's execution willingness. In addition, when referring to the user performing the training activity, the user's facial expression characteristics can be identified to represent the user's effort in performing the current training activity. This can be used as another way to evaluate the target user's adaptability or execution willingness to perform the training activity. Therefore, more features that are beneficial to the target user's adjustment basis can be mined to facilitate more personalized adjustment of the rehabilitation plan.
[0083] As an exemplary embodiment, the actual execution degree sequence of staged rehabilitation measures and / or the actual rehabilitation result sequence of stages are input into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user, which also includes: extracting the rehabilitation static feature sequence of the actual execution degree sequence of staged rehabilitation measures and / or the actual rehabilitation result sequence of stages; determining a preset rehabilitation feature sequence that is time-aligned with the rehabilitation static feature sequence based on the preset rehabilitation time node sequence in which the rehabilitation static feature sequence is located; constructing a rehabilitation difference feature sequence based on the rehabilitation static feature sequence and the preset rehabilitation feature sequence; and inputting the rehabilitation difference feature sequence into the pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user.
[0084] In this embodiment, the rehabilitation static feature sequence can be the actual rehabilitation results of the target user in the current rehabilitation stage and before the current rehabilitation stage, which are extracted from the measured follow-up data; the preset rehabilitation feature sequence that is time-aligned with the rehabilitation node corresponding to the current actual rehabilitation result is time-aligned; wherein, the preset rehabilitation feature can be obtained through the stage actual rehabilitation results of the current rehabilitation node corresponding to the historical initial rehabilitation plan and historical user portrait in the first historical rehabilitation data set, whose similarity with the target user's initial rehabilitation plan and user portrait is greater than the preset degree.
[0085] In this embodiment, the preset rehabilitation plan adjustment model is used to output the final rehabilitation plan for the target user based on the rehabilitation static feature sequence and the preset rehabilitation features; this process can output the final rehabilitation plan through the difference between the rehabilitation static feature sequence and the preset rehabilitation features. The final rehabilitation plan can make the expected rehabilitation static feature sequence of the rehabilitation time node after the preset rehabilitation time node approach the form of the stage-by-stage actual rehabilitation result corresponding to the rehabilitation time node.
[0086] As a possible implementation method, after obtaining the final rehabilitation plan based on the static rehabilitation features, the final rehabilitation plan can be integrated with the final rehabilitation plan obtained based on the time series cumulative features to obtain a new rehabilitation plan.
[0087] As an exemplary embodiment, the method also includes: when the actual rehabilitation status reaches the warning state, the measured user portrait data, the measured follow-up feedback data, and the initial rehabilitation plan are input into a pre-trained warning prediction model to obtain warning prompt information, and the warning prompt information is used to warn the target user.
[0088] In this embodiment, the preset rehabilitation outcome prediction model is trained based on the second historical rehabilitation data set; specifically, during the training process, the historical user portrait data, historical initial rehabilitation plan data, and historical follow-up data in the second historical rehabilitation data set are used as input, and the historical rehabilitation outcome data are used as the target value for model training to obtain the preset rehabilitation outcome prediction model.
[0089] After the model training is completed, the measured user portrait data, measured follow-up feedback data, and the initial rehabilitation plan are input into the pre-trained preset rehabilitation result prediction model to obtain the predicted rehabilitation result; among them, the preset rehabilitation result means that if the user continues to execute the initial rehabilitation plan according to the current rehabilitation plan execution status, it will cause problems in the part to be rehabilitated. Therefore, the target user is warned based on the predicted rehabilitation result.
[0090] As a possible implementation method, target users can be warned based on the predicted rehabilitation results by means of message push and SMS reminders.
[0091] As an exemplary embodiment, the method for constructing a preset rehabilitation plan adjustment model includes:
[0092] Multiple sets of historical user portrait data, historical initial rehabilitation plan data, historical follow-up feedback data, historical rehabilitation plan adjustment data and historical rehabilitation status of the target users are obtained as the initial data set.
[0093] The initial data set is divided based on the historical rehabilitation status to obtain a first historical rehabilitation data set; wherein the historical rehabilitation status in the first historical rehabilitation data set meets the first preset rehabilitation status.
[0094] The temporal dependency relationship of the historical follow-up feedback data is extracted to obtain the historical first temporal change characteristics and the historical second temporal change characteristics; among them, the historical first temporal change characteristics are used to represent the historical implementation degree of the rehabilitation measures, and the historical second temporal change characteristics are used to represent the phased historical results of the rehabilitation plan of the target user.
[0095] In this embodiment, the extraction of temporal dependency relationships of historical follow-up feedback data can be referred to the content in the above embodiment, which will not be repeated here.
[0096] The historical first time series change feature and the historical second time series change feature are integrated to obtain the historical time series cumulative feature; wherein the time series cumulative feature is used to represent the actual recovery status of the target user.
[0097] In this embodiment, the historical first time series change feature and the historical second time series change feature are integrated to obtain the historical time series cumulative feature. The contents in the above embodiment can be referred to and will not be repeated here.
[0098] The historical first time series change characteristics, the historical second time series change characteristics, the historical time series cumulative characteristics and the corresponding historical rehabilitation plan adjustment data are input into the initial rehabilitation plan adjustment model, and the model is trained with the historical rehabilitation plan adjustment data as the target value to obtain the preset rehabilitation plan adjustment model.
[0099] In the above-mentioned embodiment, in the process of training the preset rehabilitation plan adjustment model, by extracting the historical first time series change feature and the historical second time series change feature and fusing them to obtain the time series accumulation feature, it is possible to consider the historical data in each rehabilitation stage, the rehabilitation measures of the rehabilitation stage and the rehabilitation measures before the rehabilitation stage, the accumulation over time of the execution degree determines the actual rehabilitation result of the stage in the rehabilitation stage, the accumulation over time of all the stage rehabilitation measures currently performed by the target user, and the accumulation of the actual rehabilitation results of the stage determines the final rehabilitation result of the target user to train the preset rehabilitation plan adjustment model; therefore, in this embodiment, the actual execution of the stage rehabilitation measures is determined. The degree sequence and / or the stage-by-stage actual rehabilitation result sequence are input into a pre-trained preset rehabilitation plan adjustment model to obtain the target user's final rehabilitation plan. This step can adaptively adjust the rehabilitation plan based on the pre-trained rehabilitation plan adjustment model taking into account the target user's cumulative execution degree of the rehabilitation measures of the rehabilitation stage and the rehabilitation measures before the rehabilitation stage in each stage reflected in the measured follow-up feedback data, as well as the cumulative execution degree of all stage-by-stage rehabilitation measures currently performed by the target user over time, to obtain the target user's final rehabilitation plan that is adjusted to adapt to the training differences when the target user undergoes rehabilitation training based on the initial rehabilitation plan.
[0100] In the second aspect, this embodiment provides a rehabilitation plan adjustment device based on user portraits, such as Figure 3 Shown, including:
[0101] The first acquisition module 501 is used to obtain the measured user portrait data of the target user;
[0102] Determination module 502, for determining the initial rehabilitation plan and target follow-up content for the target user based on the measured user portrait data;
[0103] The second acquisition module 503 is used to obtain the measured follow-up feedback data of the target user in response to the target follow-up content feedback;
[0104] The rehabilitation status determination module 504 is configured to determine the target user's actual rehabilitation status based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series sequence of the actual implementation degree of each staged rehabilitation measure and a corresponding sequence of actual rehabilitation results for each stage;
[0105] The rehabilitation plan generation module 505 is used to input the actual execution degree sequence of the phased rehabilitation measures and / or the actual rehabilitation result sequence of the phased rehabilitation measures into the pre-trained preset rehabilitation plan adjustment model when the actual rehabilitation status does not reach the warning state, so as to obtain the final rehabilitation plan for the target user.
[0106] It should be noted here that the examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the contents disclosed in the above embodiments.
[0107] It should be noted that the above modules as part of the device can be implemented through software or hardware, wherein the hardware environment includes a network environment.
[0108] An embodiment of the present invention also provides a computer device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus, the memory is used to store computer programs; the processor is used to execute the method in any of the above embodiments by running the computer program stored in the memory.
[0109] Figure 4 is a structural block diagram of an optional computer device according to an embodiment of the present application, such as Figure 4 As shown, it includes a processor 10, a communication interface 20, a memory 30 and a communication bus 40, wherein the processor 10, the communication interface 20 and the memory 30 communicate with each other through the communication bus 40, wherein,
[0110] Memory 30, for storing computer programs;
[0111] The processor 10 is configured to implement the method for constructing a disease database according to any of the above embodiments when executing the computer program stored in the memory 30 .
[0112] Optionally, in this embodiment, the communication bus may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.
[0113] The communication interface is used for communication between the above-mentioned computer device and other devices.
[0114] The memory may include RAM, or may include non-volatile memory, such as at least one disk memory. Alternatively, the memory may also be at least one storage device located away from the aforementioned processor.
[0115] The above-mentioned processor can be a general-purpose processor, including but not limited to: CPU (Central Processing Unit), NP (Network Processor), etc.; it can also be DSP (Digital Signal Processing), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
[0116] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment will not be described in detail here.
[0117] It can be understood by those skilled in the art that Figure 4 The structure shown is for illustration only. The device for implementing any one of the methods in the above embodiments may be a terminal device, which may be a smart phone (such as an Android phone, an IOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (MID), a PAD, or other terminal devices. Figure 4 It does not limit the structure of the above electronic device. For example, the terminal device may also include Figure 4 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 4 Different configurations shown.
[0118] A person skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which can include: a flash drive, ROM, RAM, a magnetic disk or an optical disk, etc.
[0119] As an exemplary embodiment, the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute any one of the method steps of the present embodiment when run.
[0120] Optionally, in this embodiment, the above-mentioned storage medium can be used to execute the program code of the method steps of the embodiment of the present application.
[0121] Optionally, in this embodiment, the above-mentioned storage medium may be located on at least one network device among the multiple network devices in the network shown in the above-mentioned embodiment.
[0122] Optionally, in this embodiment, the storage medium is configured to store data for executing the method in the above embodiment.
[0123] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, which will not be described in detail in this embodiment.
[0124] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a ROM, a RAM, a mobile hard disk, a magnetic disk, or an optical disk.
[0125] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0126] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above-mentioned computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the method in the above embodiments.
[0127] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0128] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is merely a logical functional division. In actual implementation, there may be other division methods, such as combining or integrating multiple units or components into another system, or ignoring or not implementing some features. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interface, indirect coupling or communication connection of units or modules, and may be electrical or other forms.
[0129] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the purpose of the solution provided in this embodiment.
[0130] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.
[0131] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
[0132] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0133] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A method for adjusting a rehabilitation plan based on user portraits, characterized in that: The rehabilitation plan adjustment method based on user portrait includes: Acquire measured user portrait data of the target user; wherein the measured user portrait data is constructed based on the hospitalization cases and treatment data of the target user; Determine the initial rehabilitation plan and target follow-up content of the target user based on the measured user portrait data; Acquiring measured follow-up feedback data of the target user in response to the target follow-up content feedback; Determining the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of actual implementation degrees of staged rehabilitation measures and a corresponding sequence of staged actual rehabilitation results; When the actual rehabilitation status does not reach the warning state, the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence are input into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user; including: extracting the time series dependency relationship of the actual execution degree sequence of the staged rehabilitation measures to obtain a first time series change feature; the first time series change feature is used to represent the dynamic characteristics of the actual execution degree of the rehabilitation measures between different stages of the rehabilitation plan; extracting the time series dependency relationship of the staged actual rehabilitation result sequence to obtain a second time series change feature; the second time series change feature is used to represent the dynamic characteristics of the actual rehabilitation results between different stages of the rehabilitation plan; fusing the first time series change feature The first time series variation feature and the second time series variation feature are combined to obtain a time series accumulation feature; wherein the time series accumulation feature is used to represent the actual rehabilitation status of the target user; the first time series variation feature, the second time series variation feature and the time series accumulation feature are input into the preset rehabilitation plan adjustment model to obtain the final rehabilitation plan of the target user; wherein the method for extracting the first time series variation feature and the second time series variation feature comprises: determining the part to be rehabilitated of the target user based on the measured user portrait; extracting the measured follow-up feedback data based on the part to be rehabilitated to obtain the rehabilitation part feature of the part to be rehabilitated; extracting the time series dependency relationship of the rehabilitation part feature to obtain the first time series variation feature and the second time series variation feature; Wherein, the step of inputting the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user also includes: extracting the rehabilitation static feature sequence of the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence; determining a preset rehabilitation feature sequence that is aligned in time with the rehabilitation static feature sequence based on a preset rehabilitation time node sequence in which the rehabilitation static feature sequence is located; constructing a rehabilitation difference feature sequence based on the rehabilitation static feature sequence and the preset rehabilitation feature sequence; and inputting the rehabilitation difference feature sequence into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user; The method for constructing the preset rehabilitation plan adjustment model includes: obtaining multiple groups of historical user portrait data, historical initial rehabilitation plan data, historical follow-up feedback data, historical rehabilitation plan adjustment data and historical rehabilitation status of the target user as an initial data set; Dividing the initial data set based on the historical rehabilitation status to obtain a first historical rehabilitation data set; wherein the historical rehabilitation status in the first historical rehabilitation data set does not reach a warning state; Extracting the temporal dependency relationship of the historical follow-up feedback data to obtain a historical first temporal change feature and a historical second temporal change feature; wherein the historical first temporal change feature is used to represent the historical execution degree of the rehabilitation measures, and the historical second temporal change feature is used to represent the staged historical results of the rehabilitation plan of the target user; The historical first time series change feature and the historical second time series change feature are integrated to obtain a historical time series cumulative feature; wherein the time series cumulative feature is used to represent the actual rehabilitation status of the target user; The historical first time series change feature, the historical second time series change feature, the historical time series accumulation feature and the corresponding historical rehabilitation plan adjustment data are input into the initial rehabilitation plan adjustment model, and the model is trained with the historical rehabilitation plan adjustment data as the target value to obtain the preset rehabilitation plan adjustment model.
2. The method for adjusting a rehabilitation plan based on user portrait according to claim 1, characterized in that: The method for extracting the first time series change feature and the second time series change feature also includes: Extracting training activity features representing the user's rehabilitation training activity status from each of the measured follow-up feedback data; Determining the actual rehabilitation training activity state of the target user based on the training activity characteristics; wherein the actual rehabilitation training activity state includes at least one of exercise frequency, intensity, and duration; The time series dependency relationship of the actual rehabilitation training activity state is extracted to obtain the first time series change feature and the second time series change feature.
3. The method for adjusting a rehabilitation plan based on user portrait according to claim 1, characterized in that: The method further comprises: When the actual rehabilitation status reaches the warning state, the measured user portrait data, the measured follow-up feedback data, and the initial rehabilitation plan are input into a pre-trained warning prediction model to obtain warning prompt information, and the warning prompt information is used to warn the target user.
4. A rehabilitation plan adjustment device based on user portrait, characterized in that: The rehabilitation plan adjustment device based on user portrait includes: The first acquisition module is used to obtain the measured user portrait data of the target user; A determination module, used to determine the initial rehabilitation plan and target follow-up content of the target user based on the measured user portrait data; A second acquisition module is used to acquire the measured follow-up feedback data of the target user in response to the target follow-up content feedback; A rehabilitation status determination module, used to determine the actual rehabilitation status of the target user based on the initial rehabilitation plan and the measured follow-up feedback data; the measured follow-up feedback data includes a time-series-based sequence of actual execution degrees of staged rehabilitation measures and a corresponding sequence of staged actual rehabilitation results; A rehabilitation plan generation module is used to input the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence into a pre-trained preset rehabilitation plan adjustment model when the actual rehabilitation status has not reached the warning state, so as to obtain the final rehabilitation plan for the target user, including: extracting the time series dependency relationship of the actual execution degree sequence of the staged rehabilitation measures to obtain a first time series change feature; the first time series change feature is used to represent the dynamic characteristics of the actual execution degree of the rehabilitation measures between different stages of the rehabilitation plan; extracting the time series dependency relationship of the staged actual rehabilitation result sequence to obtain a second time series change feature; the second time series change feature is used to represent the dynamic characteristics of the actual rehabilitation results between different stages of the rehabilitation plan; fusing the The first time series variation feature and the second time series variation feature are used to obtain a time series accumulation feature; wherein the time series accumulation feature is used to represent the actual rehabilitation status of the target user; the first time series variation feature, the second time series variation feature and the time series accumulation feature are input into the preset rehabilitation plan adjustment model to obtain the final rehabilitation plan of the target user; wherein the method for extracting the first time series variation feature and the second time series variation feature comprises: determining the part to be rehabilitated of the target user based on the measured user portrait; extracting the measured follow-up feedback data based on the part to be rehabilitated to obtain the rehabilitation part feature of the part to be rehabilitated; extracting the time series dependency relationship of the rehabilitation part feature to obtain the first time series variation feature and the second time series variation feature; Wherein, the step of inputting the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user further includes: extracting a rehabilitation static feature sequence of the actual execution degree sequence of the staged rehabilitation measures and / or the actual staged rehabilitation result sequence; determining a preset rehabilitation feature sequence that is time-series aligned with the rehabilitation static feature sequence based on a preset rehabilitation time node sequence in which the rehabilitation static feature sequence is located; constructing a rehabilitation difference feature sequence based on the rehabilitation static feature sequence and the preset rehabilitation feature sequence; and inputting the rehabilitation difference feature sequence into a pre-trained preset rehabilitation plan adjustment model to obtain the final rehabilitation plan for the target user; The method for constructing the preset rehabilitation plan adjustment model includes: obtaining multiple groups of historical user portrait data, historical initial rehabilitation plan data, historical follow-up feedback data, historical rehabilitation plan adjustment data and historical rehabilitation status of the target user as an initial data set; Dividing the initial data set based on the historical rehabilitation status to obtain a first historical rehabilitation data set; wherein the historical rehabilitation status in the first historical rehabilitation data set does not reach a warning state; Extracting the temporal dependency relationship of the historical follow-up feedback data to obtain a historical first temporal change feature and a historical second temporal change feature; wherein the historical first temporal change feature is used to represent the historical execution degree of the rehabilitation measures, and the historical second temporal change feature is used to represent the staged historical results of the rehabilitation plan of the target user; The historical first time series change feature and the historical second time series change feature are integrated to obtain a historical time series cumulative feature; wherein the time series cumulative feature is used to represent the actual rehabilitation status of the target user; The historical first time series change feature, the historical second time series change feature, the historical time series accumulation feature and the corresponding historical rehabilitation plan adjustment data are input into the initial rehabilitation plan adjustment model, and the model is trained with the historical rehabilitation plan adjustment data as the target value to obtain the preset rehabilitation plan adjustment model.
5. A computer device, characterized in that: include: The memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the user portrait-based rehabilitation plan adjustment method of any one of claims 1 to 3 by executing the computer instructions.
6. A computer-readable storage medium, characterized in that: Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to enable a computer to execute the user portrait-based rehabilitation plan adjustment method of any one of claims 1 to 3.
Citation Information
Patent Citations
Patient follow-up visit reminding cycle self-adaptive adjustment method
CN103310109A
Pharmacy management method, server and computer storage medium
CN109493939A
Medication scheme adjustment method and device, computer equipment and storage medium
CN113689930A
User management method and device, electronic equipment and computer readable medium
CN114582453A
Exercise plan adjusting method and electronic equipment
CN114783562A
