Research On Wearable Gait Monitoring And Analysis Technology For Postoperative Rehabilitation

The gait contains a lot of human's kinematics information,which is affected by the nervous and muscular system.In modern medicine,gait is often used to assess the patient's postoperative rehabilitation,and abnormal gait is also used as a warning for some diseases.This thesis studies the wearable gait monitoring and analysis technology for postoperative rehabilitation.By attaching multiple wearable inertial sensors on the lower limbs of the human body,gait rehabilitation exercises can be monitored and analyzed.Based on this technology,we design and implement a remote gait monitoring and analysis system,including a group of wearable devices and support system software.The main content of this thesis can be summarized as follows:(1)This thesis studies the problem of space synchronization of multiple inertial sensors in gait monitoring and analysis.Since the measurements of sensors are based on the device coordinate system,the coordinate systems of each sensor at different time are different from each other,and the coordinate systems of different sensors are also different from each other.Space synchronization is a prerequisite for gait monitoring and analysis.First,this thesis designs an adaptive complementary filtering algorithm to synchronize the coordinate systems of a sensor at different time to avoid high-frequency jitter and drift errors caused by sensor defects.Then,according to the data fluctuation characteristics of acceleration and angular velocity during walking,the principal component analysis algorithm is used to extract the same forward direction and lateral direction of all sensors to achieve the synchronization of the coordinate systems of different sensors.(2)This thesis studies the correlation between the length of the lower limbs and the inertial sensor data during walking,and accurately estimates the length of the lowerlimbs without prior knowledge and training data.Existing research usually uses the length of the lower limbs to track gait movements and calculate parameters,which needs to be provided by the user in advance,or requires the user to provide enough training data.In this thesis,the geometric model of the human body and multi-dimensional sensor data are used to calculate the user's lower limb length online during gait rehabilitation exercise.No prior knowledge or additional training data is required,which makes the user more convenient when using the system.(3)This thesis designs a complete gait monitoring and analysis algorithm based on the coordinate system synchronization and estimation of the lower limb length.We use the zero-velocity calibration and the constraint of hip joints to accurately calculate the displacement of sensors;And we derive the rotation angles of sensors from the rotation matrix of synchronous coordinate systems.Thus,gait monitoring is realized.In addition,the gait parameters are calculated through the fusion of multi-dimensional motion data of multiple sensors.This thesis evaluates the accuracy of the algorithm in various scenarios.The experimental results show that,compared with the highprecision video motion capture system,the error of this algorithm for angle estimation is about 3 degrees,the error of displacement calculation is about 2.3%,and the parameter estimation error is also about 3%.(4)This thesis designs and implements a remote gait monitoring and analysis system,including a set of inexpensive wearable devices and a matching software system,patients can carry wearable devices for gait rehabilitation exercise at home,the mobile phone applications sends the sensor data to the server,and the medical staff can see the recovery animation of the gait movement and the dynamic curve of the gait parameters through the web pages.The system has been piloted in the Nanjing Military Region General Hospital and has been evaluated on different patients.The results show that the recovery animation can accurately recover the patient's gait,and the results of gait analysis are consistent with the clinical observation,which can assist the medical staff to find the abnormalities in the gait.