Gait Phase Recognition Using Deep Convolutional Neural Network with Inertial Measurement Units.

Binbin Su,Christian Smith,Elena Gutierrez Farewik

doi:10.3390/bios10090109

Binbin Su, Christian Smith + Show 1 more

Open Access

https://doi.org/10.3390/bios10090109

Copy DOI

Abstract

Gait phase recognition is of great importance in the development of assistance-as-needed robotic devices, such as exoskeletons. In order for a powered exoskeleton with phase-based control to determine and provide proper assistance to the wearer during gait, the user’s current gait phase must first be identified accurately. Gait phase recognition can potentially be achieved through input from wearable sensors. Deep convolutional neural networks (DCNN) is a machine learning approach that is widely used in image recognition. User kinematics, measured from inertial measurement unit (IMU) output, can be considered as an ‘image’ since it exhibits some local ‘spatial’ pattern when the sensor data is arranged in sequence. We propose a specialized DCNN to distinguish five phases in a gait cycle, based on IMU data and classified with foot switch information. The DCNN showed approximately 97% accuracy during an offline evaluation of gait phase recognition. Accuracy was highest in the swing phase and lowest in terminal stance.

Highlights

Exoskeletons are mechanical devices that can help augment a person’s strength or assist movement in people with motor disorders
The Deep convolutional neural networks (DCNN) detected Sw with the highest classification accuracy (99.3%), followed by PSw (96.2%), MS (95.8%) and loading response (LR) (95.8%)
Gait is a periodic movement with a repetitive nature and temporal pattern, we explored in this study the treatment of gait phases as discrete events in which strong correlations may exist among measured sensor data within the same gait phases, as well as large difference between different gait phases

Summary

Introduction

Exoskeletons are mechanical devices that can help augment a person’s strength or assist movement in people with motor disorders. A gait cycle can be divided into stance when the foot is in contact with the ground and swing when the limb has no contact with the ground, and further demarcated by gait events, such as ispilateral foot contact, contralateral foot off, etc. Accurate identification of gait phases is crucial for metabolically-efficient control of lower limb exoskeletons; inaccurately detected gait phases tend to either increase the user’s effort or exert improper joint torque [4]. When a patient wears an exoskeleton, a gait phase-modulated torque is commonly provided to assist the patient. Martini et al [5] described hip flexor assistance during swing to reduce the total energy cost of walking with a powered hip exoskeleton called the Active Pelvis

Objectives

Results

Discussion

Conclusion