CN-121983240-A - Leg fracture bedridden full-period rehabilitation system based on mixed reality haptic interaction and control method

CN121983240ACN 121983240 ACN121983240 ACN 121983240ACN-121983240-A

Abstract

The invention relates to the technical field of medical rehabilitation equipment control, in particular to a leg fracture bedridden full-period rehabilitation system and a control method based on mixed reality haptic interaction, which are adopted to construct a multi-mode rehabilitation state feature set, form a coupling feature set through feature coupling, dynamically update a local dynamic transition threshold value and a global dynamic transition threshold value, set triggering conditions, judge whether to transition from a current rehabilitation stage to a next stage according to the triggering conditions, and dynamically updating the local dynamic transition threshold and the global accumulation threshold, and then sending a control signal for virtual display to the MR interaction unit according to the current rehabilitation stage and preset conditions, so as to realize real-time self-adaptive switching of multi-stage rehabilitation training, realize the function of mixed reality by combining with the MR interaction unit, provide light, immersive and personalized rehabilitation experience for patients, accelerate the recovery of leg functions and improve the rehabilitation efficiency and quality.

Inventors

WANG YONGHUA
WANG SHAN
HOU XUELIN
LIU DAN
LIANG RUIPING
WU XINGYU
GUO ZIHU

Assignees

陕西九为云智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260210

Claims (10)

1. The leg fracture bedridden full-cycle rehabilitation system control method based on mixed reality haptic interaction is characterized by comprising the following steps of: Acquiring multi-modal data of a current patient, and performing data preprocessing on the multi-modal data; respectively extracting features of the multi-mode data after data preprocessing, carrying out feature fusion on the obtained features, constructing a multi-mode rehabilitation state feature set, and forming a coupling feature set through feature coupling; The method comprises the steps of constructing a self-adaptive neural network model, wherein the self-adaptive neural network model takes a multi-mode feature set and a coupling feature set as input, outputs a coupling step value in a rehabilitation stage and a cumulative coupling comprehensive rehabilitation level, constructs a local dynamic transition threshold based on the coupling step value in the rehabilitation stage, and constructs a global dynamic transition threshold based on the cumulative coupling comprehensive rehabilitation level; Optimizing model parameters through a loss function, dynamically updating a local dynamic transition threshold and a global dynamic transition threshold, setting a trigger condition, and judging whether to transition from a current rehabilitation stage to a next stage according to the trigger condition; And constructing a plurality of rehabilitation training scenes based on different rehabilitation stages, and automatically triggering self-adaptive switching of the rehabilitation training scenes based on the updated local dynamic transition threshold and the updated global dynamic transition threshold.
2. The leg fracture bedridden full-cycle rehabilitation system control method based on mixed reality haptic interaction of claim 1, wherein the adaptive neural network model comprises an input layer, a multi-stage dynamic feature weight branching layer, a coupling feature interaction layer, a tired period memory layer, a threshold learning layer and an output layer: The input layer is used for receiving a set of multi-mode characteristic data which can be acquired by a patient currently; The coupling feature interaction layer is used for receiving a standardized coupling feature set, quantifying coupling feature interaction strength through a multi-head attention mechanism and a full connection layer and converting the coupling feature interaction strength into a coupling effect coefficient; the accumulation memory fusion layer is used for balancing the contribution of the history memory and the current state through the learnable history phase contribution attenuation coefficient, The threshold parameter learning layer is used for dynamically adjusting the transition threshold by learning the threshold parameter; the output layer substitutes the calculated coupling step value of the rehabilitation stage, the local dynamic transition threshold value, the integrated rehabilitation level of the cumulative coupling and the global dynamic transition threshold value into the triggering conditions, when the conditions are all met, 1 is output, the transition is triggered, and otherwise 0 is output.
3. The leg fracture bedridden full cycle rehabilitation system control method based on mixed reality haptic interaction according to claim 2, wherein the triggering condition comprises: A first condition that data acquisition of a next stage feature can be performed based on a current patient state; Second condition: In the formula, For the rehabilitation phase to couple the step value, For the local dynamic transition threshold value, For the integrated rehabilitation level of the coupling in the tired period, Is a global dynamic transition threshold; When the first condition and the second condition are satisfied simultaneously, a transition is made from the current rehabilitation stage to the next stage.
4. The leg fracture bedridden full-cycle rehabilitation system control method based on mixed reality haptic interaction according to claim 1, wherein constructing a plurality of rehabilitation training scenes based on different rehabilitation stages comprises: the rehabilitation stage of the current patient is obtained, preset conditions of a plurality of rehabilitation stages are respectively set, and a control signal for virtual display is sent to the MR interaction unit according to the current rehabilitation stage and the corresponding preset conditions.
5. The method for controlling a full-period rehabilitation system for leg fracture and bedridden patients based on mixed reality haptic interaction according to claim 1, wherein the multi-modal data comprises basic physiological data, bone structure and morphology data, biomechanical data, functional and movement data and feedback data.
6. The method for controlling a full-period rehabilitation system for leg fracture and bed according to claim 1, wherein the rehabilitation stages comprise a bed rest period, an off-bed transition period and an autonomous walking preparation period; Acquiring a starting time node of the end of the operation of the patient, acquiring a current target time node, acquiring current postoperative time according to the starting time node and the target time node, and outputting the type of the rehabilitation stage currently judged based on the postoperative time.
7. The method for controlling a full-period rehabilitation system of leg fracture and bedridden based on mixed reality haptic interaction according to claim 4, further comprising projecting a three-dimensional virtual skeleton model of a lower limb of a patient side if a current rehabilitation stage is a bedridden rest stage, setting a safe activity area, and distinguishing and displaying with colors different from the three-dimensional virtual skeleton model; setting an alarm threshold and a safety threshold, and sending out prompt and control signals with different colors based on the maximum hip joint movement angle, the alarm threshold and the safety threshold; When a control signal for performing the first training is received, projecting a virtual track moving at a set speed, collecting the angular fluctuation rate of the current patient in real time, setting a first angular fluctuation rate threshold, and when the angular fluctuation rate is greater than the first angular fluctuation rate threshold, transmitting a control signal for stopping projecting the virtual track and starting an air bag feedback assembly.
8. The method for controlling a full-period rehabilitation system for leg fracture and bedridden patients based on mixed reality haptic interaction according to claim 7, further comprising the steps of, if the current rehabilitation stage is an out-of-bed transition stage, the preset condition comprising a first judgment threshold and a second judgment threshold, the first judgment threshold being greater than the second judgment threshold, further comprising: when a control signal for performing the second training is received, a plurality of virtual barriers with the height of N meters are projected, the maximum activity angle of the hip joint during the training of the patient at the moment is obtained, and if the maximum activity angle is smaller than a first judgment threshold value, the height of the virtual barriers is adjusted through a height adjustment model to be redisplayed; If the maximum activity angle of the hip joint is smaller than the second judgment threshold value, sending a notification signal for suspending the second training, and after the first training is needed again, carrying out the second training.
9. The method for controlling a full-cycle rehabilitation system for leg fracture and bedridden based on mixed reality haptic interaction according to claim 8, further comprising if the current rehabilitation phase is an autonomous walking preparation phase, wherein the preset condition includes a step judgment threshold value obtained based on a maximum activity angle of a hip joint, further comprising: According to the initial step length setting, projecting a virtual footprint which the patient needs to advance to next step according to the current patient position, and obtaining the actual step length of the patient according to the acquired maximum activity angle and the step time of the hip joint; setting a second angle fluctuation rate, and when the angle fluctuation rate is larger than the second angle fluctuation rate, sending out standard demonstration actions for display; Generating a virtual step, setting a proportional relation between the height of the virtual step and the maximum activity angle of the hip joint, and adjusting the height of the virtual step based on the numerical value of the maximum activity angle of the hip joint; and adjusting the resistance value of the SMA wire bundles according to the height of the virtual steps.
10. Leg fracture bedridden full cycle rehabilitation system based on mixed reality haptic interaction is characterized by comprising: The wearable device comprises an outer skin-friendly layer, a middle layer sensing layer and an inner layer feedback layer which are sequentially arranged, wherein a pressure sensor, a triaxial accelerometer and an inertial measurement unit are arranged in the middle layer sensing layer and are used for collecting data, and the inner layer feedback layer is provided with an SMA wire bundle and an air bubble feedback assembly; The MR display device is used for receiving the control signal to display the virtual picture; A master control device connected with the wearable device and the MR display apparatus for executing a leg fracture bedridden full cycle rehabilitation system control method based on hybrid realistic haptic interaction as claimed in any one of claims 1-9.

Description

Leg fracture bedridden full-period rehabilitation system based on mixed reality haptic interaction and control method Technical Field The invention relates to the technical field of medical rehabilitation equipment control, in particular to a leg fracture bedridden full-period rehabilitation system based on mixed reality haptic interaction and a control method. Background In the prior art, the traditional rehabilitation equipment mainly uses mechanical movement, only provides single passive movement of limbs, and lacks multi-mode feedback such as vision, touch sense and the like. The virtual environment constructed by the existing VR rehabilitation system is independent of the real physical space, and the action instructions in the virtual scene cannot be accurately matched with the real limb movement state of the patient. For example, the height of the leg is required to be 20cm in the virtual scene, but when the actual leg lifting height of a patient is insufficient, the device cannot adjust feedback in real time, so that training actions deform, a compensatory movement mode can be formed after long-term use, the rehabilitation effect is affected, and even secondary damage can be caused. The training parameters (such as resistance and movement amplitude) of the existing rehabilitation equipment cannot be automatically adjusted according to the real-time state (such as muscle fatigue degree and joint movement degree change) in the training process of a patient. Once the patient is tired or uncomfortable in training, if the medical staff does not find in time, the risk of training interruption or damage is possibly increased, and rehabilitation continuity and safety are affected. The existing rehabilitation technology mainly makes a training plan based on the current state, and does not have the prediction capability of patient rehabilitation trend. The potential rehabilitation risk (such as early signs of joint stiffness) cannot be identified in advance, and the scheme is adjusted, so that the problems are usually passively dealt with after the problems occur, the optimal intervention time is missed, the rehabilitation period is prolonged, and the pain and the medical cost of patients are increased. Disclosure of Invention The invention aims to provide a leg fracture bedridden full-period rehabilitation system based on mixed reality haptic interaction and a control method thereof, which solve the problems in the prior art. The invention is realized by the following technical scheme: In a first aspect, the invention provides a leg fracture bedridden full-cycle rehabilitation system control method based on mixed reality haptic interaction, which comprises the following steps: Acquiring multi-modal data of a current patient, and performing data preprocessing on the multi-modal data; respectively extracting features of the multi-mode data after data preprocessing, carrying out feature fusion on the obtained features, constructing a multi-mode rehabilitation state feature set, and forming a coupling feature set through feature coupling; The method comprises the steps of constructing a self-adaptive neural network model, wherein the self-adaptive neural network model takes a multi-mode feature set and a coupling feature set as input, outputs a coupling step value in a rehabilitation stage and a cumulative coupling comprehensive rehabilitation level, constructs a local dynamic transition threshold based on the coupling step value in the rehabilitation stage, and constructs a global dynamic transition threshold based on the cumulative coupling comprehensive rehabilitation level; Optimizing model parameters through a loss function, dynamically updating a local dynamic transition threshold and a global dynamic transition threshold, setting a trigger condition, and judging whether to transition from a current rehabilitation stage to a next stage according to the trigger condition; And constructing a plurality of rehabilitation training scenes based on different rehabilitation stages, and automatically triggering self-adaptive switching of the rehabilitation training scenes based on the updated local dynamic transition threshold and the updated global dynamic transition threshold. Constructing a plurality of rehabilitation training scenes based on different rehabilitation stages comprises: the rehabilitation stage of the current patient is obtained, preset conditions of a plurality of rehabilitation stages are respectively set, and a control signal for virtual display is sent to the MR interaction unit according to the current rehabilitation stage and the corresponding preset conditions. Preferably, the multimodal data includes basic physiological data, skeletal structure and morphology data, biomechanical data, functional and kinetic data, and feedback data. Preferably, the types of rehabilitation stages include a bedridden period, an off-bed transition period, and an autonomous walking preparation period; Acquiring