A knee joint rehabilitation method and system based on graphene flexible sensors

Abstract

This invention discloses a knee joint rehabilitation method and system based on a graphene flexible sensor, relating to the field of pattern recognition. The invention includes the following steps: acquiring joint activity signals of a subject, wherein the joint activity signals are collected by a graphene flexible sensor; estimating knee joint angle data based on the subject's joint activity signals using a CurXGB model; mapping the knee joint angle data one-to-one with the angles of a video game; and having the subject complete the game through joint movements. This invention proposes a CurXGB regression algorithm that can generate more continuous prediction curves, providing rehabilitation exoskeleton robots with motion curves that better conform to human joint movements. Utilizing the characteristics of the graphene flexible sensor, it is applied to knee joint rehabilitation, achieving knee joint rehabilitation training through a graphene flexible sensor-based knee joint rehabilitation system, thereby increasing subject participation.

Description

Technical Field

[0001] This invention relates to the field of pattern recognition, and more specifically to a knee joint rehabilitation method and system based on a graphene flexible sensor. Background Technology

[0002] Hemiplegia is a motor dysfunction caused by nerve damage. Clinical treatment for hemiplegic patients primarily involves one-on-one rehabilitation therapy under the guidance of a rehabilitation therapist. This approach is time-consuming and costly. Therefore, there is an urgent need for a new and efficient rehabilitation therapy to compensate for the shortcomings of traditional rehabilitation training. Rehabilitation exoskeleton robots combine robotics technology with rehabilitation medicine, freeing rehabilitation therapists from major physical labor and enabling them to develop better rehabilitation plans for patients, thereby improving rehabilitation outcomes.

[0003] Rehabilitation training can be divided into active and passive methods. Studies have shown that active rehabilitation is more effective than passive rehabilitation. Therefore, extracting the patient's movement intentions and integrating them into a rehabilitation exoskeleton robot is an effective way to improve the patient's initiative and thus achieve more efficient rehabilitation results. Summary of the Invention

[0004] In view of this, the present invention provides a knee joint rehabilitation method and system based on a graphene flexible sensor.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] One aspect discloses a knee joint rehabilitation method based on a graphene flexible sensor, including the following steps:

[0007] The joint movement signals of the subject were acquired, and the joint movement signals were collected by a graphene flexible sensor;

[0008] Knee joint angle data were estimated using the CurXGB model based on the subjects' joint activity signals;

[0009] Match the knee joint angle data with the angles in the video game;

[0010] The subjects completed the game through joint movements.

[0011] Optionally, the estimated knee joint angle can be processed by curve shaping, as shown in the following formula:

[0012]

[0013] Where x input For input data, x split For ω * The corresponding optimal split point x value; x L and x R ω* The corresponding left and right extrema of the input value, μ is the learning rate, K is the number of decision trees, and τ is the distance adjustment factor; when making predictions, when x input Greater than x split When, the output is greater than y prep Large predicted values; when x input Less than x split When, the output is greater than y prep Small predicted values; the distance adjustment factor τ is used to adjust the degree of distance. When τ is too large, the predicted values ​​will be greatly distorted, indicating overfitting.

[0014] The knee angles predicted using XGBoost are segmented and not smooth enough, failing to closely match the knee joint's motion curve. The knee angles predicted using CurXGB, on the other hand, are smooth and better match the knee joint's motion curve.

[0015] On the other hand, a knee joint rehabilitation system based on graphene flexible sensors was also disclosed, including a sensor module, a lower-level computer module and an upper-level computer module;

[0016] The sensor module is used to collect changes in knee joint movement; the lower-level computer module is used to receive signals sent by the sensor module and transmit the signals to the upper-level computer module.

[0017] The sensor module uses a graphene sensor.

[0018] Optionally, the host computer includes a processing program and a video game; the processing program is used to process the sensor signals transmitted by the slave computer and estimate the angle of the knee joint based on the sensor signals, and the angle of the knee joint corresponds one-to-one with the angle of the video game.

[0019] Optionally, the host computer may also include a screen game, which provides visual feedback to the subject.

[0020] Optionally, the host computer includes desktop computers, PCs, and embedded devices.

[0021] Optional lower-level devices include Arduino, microcontrollers, and Raspberry Pi.

[0022] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a knee joint rehabilitation method and system based on graphene flexible sensors, which has the following beneficial effects:

[0023] 1. A CurXGB regression algorithm is proposed, which can generate more continuous prediction curves and provide motion curves that are more in line with human joint movements for rehabilitation exoskeleton robots.

[0024] 2. A knee joint rehabilitation system based on graphene flexible sensors is proposed. The characteristics of graphene flexible sensors are utilized and applied to knee joint rehabilitation. Knee joint rehabilitation training can be achieved through the knee joint rehabilitation system based on graphene flexible sensors, thereby improving the participation of subjects. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0027] Figure 2 This is a flowchart of the CurXGB model training process of the present invention;

[0028] Figure 3 This is a flowchart of the CurXGB model prediction process of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] This invention discloses a knee joint rehabilitation method based on a graphene flexible sensor, such as... Figure 1 As shown, it includes the following steps:

[0031] The joint movement signals of the subjects were acquired by a graphene flexible sensor.

[0032] Knee joint angle data were estimated using the CurXGB model based on the subjects' joint activity signals;

[0033] Match the knee joint angle data with the angles in the video game;

[0034] The subjects completed the game through joint movements.

[0035] Furthermore, the estimated knee joint angle is processed by curve shaping, as shown in the following formula:

[0036]

[0037] Where x input For input data, x split For ω * The corresponding optimal split point x value; x L and x R ω * The corresponding left and right extrema of the input value, μ is the learning rate, K is the number of decision trees, and τ is the distance adjustment factor; when making predictions, when x input Greater than x split When, the output is greater than y prep Large predicted values; when x input Less than x split When, the output is greater than y prep Small predicted values; the distance adjustment factor τ is used to adjust the degree of distance. When τ is too large, the predicted values ​​will be greatly distorted, indicating overfitting.

[0038] CurXGB and XGBoost are the same during the training phase; the improvement lies in the inference phase. When using XGBoost for regression, the label values ​​are divided into segments, for example, samples 1-3 correspond to a label of 100, and 3-9 correspond to a label of 200. Therefore, during inference, sample 2 corresponds to 100, and 6 corresponds to 200. However, this method is not smooth enough. CurXGB improves by assigning 98 to sample 2, effectively curving the original 100. XGBoost assigns 100 to every value from 1 to 3; CurXGB assigns 90-100 to every pair of values ​​from 1 to 3.

[0039] like Figure 2-3 As shown, the use of the knee joint rehabilitation system based on graphene flexible sensors is further divided into training and usage phases.

[0040] On the other hand, this invention also discloses a knee joint rehabilitation system based on a graphene flexible sensor, comprising a sensor module, a lower-level computer module, and an upper-level computer module. The graphene flexible sensor is sewn into the knee brace to form the sensor module, responsible for collecting knee joint movement changes. The lower-level computer module receives the sensor electronic signals sent by the sensor module and transmits them to the upper-level computer module. The upper-level computer module mainly includes a processing program and a video game. The processing program processes the signals transmitted from the lower-level computer and estimates the knee joint angle based on the signals; the knee joint angle corresponds to the video game. The video game is a bird game where the bird continuously traverses obstacles, and the bird's ascent and descent correspond one-to-one with the knee joint angle. The upper-level computer can process the data collected by the sensor module, estimate the knee joint angle based on the data, and then control the bird's up-and-down movement.

[0041] First, the subjects wore knee braces with sewn-on graphene flexible sensors and devices capable of detecting knee joint angles, and performed corresponding knee flexion and extension movements. During this process, graphene deformation data and knee joint angle change data were collected. Then, the graphene deformation data and knee joint angle change data were input into the CurXGB model for training. CurXGB is a regression algorithm that establishes a one-to-one mapping relationship between graphene deformation data and actual knee joint angles.

[0042] Subjects wearing knee braces with sewn-on graphene flexible sensors freely perform knee flexion and extension movements. The knee rehabilitation system collects data on the changes in these graphene flexible sensors during the subjects' movements. This data is then input into a trained CurXGB model. CurXGB estimates the knee joint angle in real time based on the collected graphene flexible sensor data. This estimation process is completed on a host computer, which uses the estimated knee joint angle data to control the ascent and descent of a bird in a video game.

[0043] Furthermore, the host computer can be a desktop computer, PC, embedded device, etc., while the slave computer can be an Arduino, microcontroller, Raspberry Pi, etc. The graphene flexible sensor can be viewed as a stretched variable resistor, using a bridge circuit to convert the variable resistance into an analog voltage.

[0044] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0045] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for knee rehabilitation based on a graphene flexible sensor, characterized in that, The method comprises the following steps: Obtaining joint movement signals of a subject, which are collected by a graphene flexible sensor; Estimating knee joint angle data based on the joint movement signals of the subject through a CurXGB model; Mapping the knee joint angle data to the angles of a video game one by one; The subject completes the game through joint movement; The estimated knee joint angle is curve processed, and the formula is as follows: ； wherein is the input data, is the corresponding optimal split point, the value; and is the left and right extreme values of the corresponding input value, is the learning rate, is the number of decision trees, is the distance adjustment factor; when is greater than , a prediction value greater than is output; when is less than , a prediction value smaller than is output; the distance adjustment factor is used to adjust the degree of distance, when is too large, the prediction value will appear large deformation, is the over-fitting state.

2. A knee rehabilitation system based on graphene flexible sensor, applying the knee rehabilitation method based on graphene flexible sensor in claim 1, characterized in that, It comprises a sensor module, a lower computer module and an upper computer module; The sensor module is used to collect knee joint movement changes; the lower computer module is used to receive signals sent by the sensor module and transmit the signals to the upper computer module; The sensor module adopts a graphene sensor.

3. The graphene-based flexible sensor based knee rehabilitation system according to claim 2, wherein, The upper computer comprises a processing program and a video game; the processing program is used to process the sensor signals transmitted by the lower computer and estimate the angle of the knee joint according to the sensor signals, and the angle of the knee joint corresponds to the angle of the video game one by one.

4. The knee rehabilitation system based on graphene flexible sensor according to claim 2, wherein, The upper computer further comprises a screen game, which provides visual feedback to the subject.

5. The graphene-based flexible sensor based knee rehabilitation system according to claim 2, wherein, The upper computer comprises a desktop computer, a PC and an embedded device.

6. The graphene-based flexible sensor based knee rehabilitation system according to claim 2, wherein, The lower computer comprises an Arduino, a single-chip microcomputer and a Raspberry Pi.

Images