Research on the Improved CNN Deep Learning Method for Motion Intention Recognition of Dynamic Lower Limb Prosthesis

Abstract

Objective. In order to study the motion recognition intention of lower limb prosthesis based on the CNN deep learning algorithm. Methods. A convolutional neural network (CNN) model was established to reconstruct the motion pattern. Before the movement mode of the affected side was converted, the sensor was bound to the healthy side. The classifier was employed to extract and classify the features, so as to realize the accurate description of the movement intention of the disabled. Results. The method proposed in this research can achieve 98.2% recognition rate of the movement intention of patients with lower limb amputation under different terrains, and the recognition rate can reach 97% after the pattern converted between the five modes was added. Conclusion. The deep learning algorithm that automatically recognized and extracted features can effectively improve the control performance on the intelligent lower limb prosthesis and realize the natural and seamless conversion of the intelligent prosthesis in a variety of motion modes.

Integrating intention-based systems in human-robot interaction: a scoping review of sensors, algorithms, and trust

“… Wang (2021) proposes the use of a convolutional neural network (CNN) model to reconstruct the motion pattern of a lower limb prosthesis. The input to the CNN model is the data collected from the IMU sensor attached to the prosthesis.…”

“…The second most commonly used sensors are IMUs and sEMG sensors. IMUs are often found in exoskeleton designs and commercialized armbands, such as the Myo armband ( Ren et al, 2018 ; Cote-Allard et al, 2019 ; Su et al, 2019 ; Gardner et al, 2020 ; Viekash et al, 2021 ; Wang, 2021 ), while sEMG sensors are mainly used to measure upper-limb movements with armbands ( Cote-Allard et al, 2019 ; Young et al, 2019 ; Chen et al, 2020a ; Chen et al, 2020b ) and lower-limb movements with sEMG electrodes ( Kilic and Dogan, 2017 ; Kopke et al, 2020 ; Coker et al, 2021 ; Feleke et al, 2021 ; Viekash et al, 2021 ; Wen and Wang, 2021 ). IMUs, which measure body acceleration and angular rate, offer the advantage of being unobtrusive, portable, and relatively easy to use, making them ideal for real-time, dynamic motion tracking.…”

Adaptive Control System of Intelligent Lower Limb Prosthesis Based on 5G Virtual Reality

“…In the treatment process of rehabilitation sports training, the patient's visual and auditory perception and the body's selfperception are used to judge the effect of the training and complete the feedback on the results of the rehabilitation training. Usually, this rehabilitation process is long and boring, and patients must be fully psychologically prepared [24].…”

Visualized Analysis of Intention Recognition Research Based on CiteSpace

“…Some scientists have applied Hidden Markov Model (HMM) [7] to predict human intention. As machine learning can accurately classify intentions [8] with data provided, many related algorithms such as attention-based Long Short-Term Memory (LSTM) Network [8], and Convolutional Neural Network and Herman Neural Network (CNN-ENN) [10] have been applied into IR. Furthermore, sensors are well developed recently to recognize human motions [10], the information about human muscles [5] or movements of other vehicles [11].…”