MEMS Inertial Sensors Based Gait Analysis for Rehabilitation Assessment via Multi-Sensor Fusion

Abstract

Gait and posture are regular activities which are fully controlled by the sensorimotor cortex. In this study, fluctuations of joint angle and asymmetry of foot elevation in human walking stride records are analyzed to assess gait in healthy adults and patients affected with gait disorders. This paper aims to build a low-cost, intelligent and lightweight wearable gait analysis platform based on the emerging body sensor networks, which can be used for rehabilitation assessment of patients with gait impairments. A calibration method for accelerometer and magnetometer was proposed to deal with ubiquitous orthoronal error and magnetic disturbance. Proportional integral controller based complementary filter and error correction of gait parameters have been defined with a multi-sensor data fusion algorithm. The purpose of the current work is to investigate the effectiveness of obtained gait data in differentiating healthy subjects and patients with gait impairments. Preliminary clinical gait experiments results showed that the proposed system can be effective in auxiliary diagnosis and rehabilitation plan formulation compared to existing methods, which indicated that the proposed method has great potential as an auxiliary for medical rehabilitation assessment.

Fast Wearable Sensor–Based Foot–Ground Contact Phase Classification Using a Convolutional Neural Network with Sliding-Window Label Overlapping

“…Human motion recognition (HMR) is a technology domain that recognizes and distinguishes different types of human activities using sensor data [ 1 ]; it is widely used in rehabilitation and medical treatment like the classification and rehabilitation evaluation of patients with hip osteoarthritis, neurological disorders such as stroke, and Parkinson’s disease through gait analysis [ 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 ]. It also has been used in training assistance like exercise coaching through motion tracking and feedback, speed and position tracking in sports training [ 12 , 13 , 14 , 15 , 16 , 17 ], sudden fall prevention [ 18 ] along with the development of wearable sensor technology.…”

“…However, as their methods were based on measuring the physical contact between the foot and the ground, lower-limb motion could not be analyzed simultaneously. Qui et al [ 7 ] and Mohammad et al [ 9 ] proposed a method for the independent detection of walking, squatting, and jumping using wearable inertial sensors. Although a wearable inertial sensor is very easy to use and has limitless measurement workspace [ 21 ], acceptable detecting accuracy has not been continuously obtained owing to sensor drifts as well as initial calibration issues [ 22 ].…”

Person Re-Identification Using Deep Modeling of Temporally Correlated Inertial Motion Patterns

“…Besides this, human gait data through inertial sensors have been used to solve different problems. For instance, Qiu et al [39] differentiated between healthy and unhealthy adults using gait data obtained through inertial sensors. An inertial motion capture system is proposed in [40] that combines visual and inertial sensors to estimate human gait with low cost and high accuracy.…”

Towards Wearable-Inertial-Sensor-Based Gait Posture Evaluation for Subjects with Unbalanced Gaits

“…Note that traditional gait analysis performed in a specific gait lab is too expensive and inconvenient. It is necessary to develop an ambulatory gait analysis method without a complex process, which is helpful for physiotherapists to easily discover abnormal gaits and the causes of chronic gait pain, as well as for fitness instructors to discover exercise injury risks and to provide guidance [7][8][9].…”