Sensory Systems in Micro-Processor Controlled Prosthetic Leg: A Review

Abstract

Microprocessor controlled prosthetic legs (MPCPL) offer better functionality than conventional prosthetic legs as they use actuators to replace missing joint function. This potentially reduces the user's metabolic energy consumption and normal walking gait can be mimicked as closely as possible. However, MPCPL require a good control system to perform efficiently, and one of the essential components is the system of sensors. The sensory system must satisfy two important criteria; the practicality in donning and doffing the prosthesis, i.e. the process of putting on and taking off the prosthesis by the amputee user, and the quality in the information provided. In this paper, a comprehensive review was conducted on studies related to the state of the art of sensory system adopted in MPCPL. The publications were searched using four electronics databases within the last 13 years. A total of 31 papers were reviewed. The articles were classified into three main categories: prosthetic-device oriented, user's-biological-input oriented and neuromechanical fusion sensory system. Types of sensors used and their application to the prosthetic system were analyzed. This review indicates that the sensors technology reported in the literature still does not fulfil the criteria of an efficient sensory system. Hence, a sensory system that eases the don and doff process of the prosthesis, yet informative in terms of providing enough useful data to effectively control the prosthesis, is needed for a successful MPCPL.

Performance of Sonomyographic and Electromyographic Sensing for Continuous Estimation of Joint Torque During Ambulation on Multiple Terrains

“…Multiple researchers have evaluated the fusion of multiple sensing types for improved estimation of lower-limb motion [47]- [49]. The majority of these studies evaluate neuromechanical (i.e., EMG and mechanical sensing) fusion for classification of ambulation modes and ambulatory transitions; on average, there is a reduction in classification error to 2.3% when neuromechanical sensors were used for classification of transitions, in comparison to surface EMG sensors only (20.2%) and mechanical sensors only (7.8%) [50]. To our knowledge, one research group has evaluated the fusion of sonomyography with surface EMG data for the continuous prediction of ankle joint moment [34].…”

Design of a Bio-Inspired Gait Phase Decoder Based on Temporal Convolution Network Architecture With Contralateral Surface Electromyography Toward Hip Prosthesis Control

“…Although our architecture has followed the sensor fusion suggestion in literature (Hamzaid et al, 2020 ), and the statistical analysis results ( Figures 3B , 6 ) indicated a high bilateral correlation and spotted no significant decoding difference when the modality choice changed, the time-variant sEMG can still confuse the decoder from time to time; this calls for the further investigation in improving the motion intention estimation accuracy.…”

Electromyography-based control of lower limb prostheses: a systematic review

“…Surgical techniques such as Targeted Muscle Reinnervation (TMR) [10] or Agonist-antagonist Myoneural Interface (AMI) [11] have proven successful in overcoming this problem by creating new EMG sources for prosthetic leg control. Moreover, a variety of signal processing and machine learning algorithms have been employed to further improve the neural decoding of motor commands in individuals with lower limb amputations [12]- [17].…”