Classification Performance and Feature Space Characteristics in Individuals With Upper Limb Loss Using Sonomyography

Abstract

Objective: Sonomyography, or ultrasound-based sensing of muscle deformation, is an emerging modality for upper limb prosthesis control. Although prior studies have shown that individuals with upper limb loss can achieve successful motion classification with sonomyography, it is important to better understand the time-course over which proficiency develops. In this study, we characterized user performance during their initial and subsequent exposures to sonomyography. Method: Ultrasound images corresponding to a series of hand gestures were collected from individuals with transradial limb loss under three scenarios: during their initial exposure to sonomyography (Experiment 1), during a subsequent exposure to sonomyography where they were provided biofeedback as part of a training protocol (Experiment 2), and during testing sessions held on different days (Experiment 3). User performance was characterized by offline classification accuracy, as well as metrics describing the consistency and separability of the sonomyography signal patterns in feature space. Results: Classification accuracy was high during initial exposure to sonomyography (96.2 ± 5.9%) and did not systematically change with the provision of biofeedback or on different days. Despite this stable classification performance, some of the feature space metrics changed. Conclusions: User performance was strong upon their initial exposure to sonomyography and did not improve with subsequent exposure. Clinical Impact: Prosthetists may be able to quickly assess if a patient will be successful with sonomyography without submitting them to an extensive training protocol, leading to earlier socket fabrication and delivery.

First Demonstration of Functional Task Performance Using a Sonomyographic Prosthesis: A Case Study

“…For example, we previously used SMG to identify five individual digit movements in able-bodied individuals with 97% cross-validation accuracy ( Sikdar et al, 2014 ) and fifteen complex hand grasps with 91% cross-validation accuracy ( Akhlaghi et al, 2016 ). We also found that, with minimal training required, SMG can identify five grasps for individuals with upper limb loss with 96% cross-validation accuracy ( Dhawan et al, 2019 ; Engdahl et al, 2022 ). Thus, it is not surprising that SMG is becoming a promising option for hand gesture recognition and prosthesis control for able-bodied individuals ( Chen et al, 2010 ; Shi et al, 2010 ; Yang et al, 2019 , 2020 ) and individuals with upper limb loss ( Zheng et al, 2006 ; Hettiarachchi et al, 2015 ; Baker et al, 2016 ; Dhawan et al, 2019 ).…”

“…It is also a limitation that our participant had congenital limb absence, as this restricted the number of distinct muscle contraction patterns she was able to produce (i.e., wrist flexion, wrist extension, rest). Our prior work has shown that many individuals with amputation can produce a higher number of distinct muscle contraction patterns corresponding to different hand gestures (rather than just wrist flexion and extension) and that these classes were successfully identified in offline testing using SMG ( Dhawan et al, 2019 ; Engdahl et al, 2022 ). Future work should explore whether real-time classification performance remains accurate when an increased number of classified hand grasps are included.…”

First-in-human demonstration of functional task performance using a sonomyographic prosthesis

“…For example, we previously used SMG to identify ve individual digit movements in able-bodied individuals with 97% cross-validation accuracy (Sikdar et al, 2014) and fteen complex hand grasps with 91% cross-validation accuracy (Akhlaghi et al, 2016). We also found that, with minimal training required, SMG can identify ve grasps for individuals with upper limb loss with 96% cross-validation accuracy (Dhawan et al, 2019;Engdahl et al, 2022). Thus, it is not surprising that SMG is becoming a promising option for hand gesture recognition and prosthesis control for able-bodied individuals (Chen et al, 2010;Shi et al, 2010;Yang et al, 2019Yang et al, , 2020 and individuals with upper limb loss (Zheng et al, 2006;Hettiarachchi et al, 2015;Baker et al, 2016;Dhawan et al, 2019).…”

“…It is also a limitation that our participant had congenital limb absence, as this restricted the number of distinct muscle contraction patterns she was able to produce (i.e., wrist exion, wrist extension, rest). Our prior work has shown that many individuals with amputation can produce a higher number of distinct muscle contraction patterns corresponding to different hand gestures (rather than just wrist exion and extension) and that these classes were successfully identi ed in o ine testing using SMG (Dhawan et al, 2019;Engdahl et al, 2022). Future work should explore whether real-time classi cation performance remains accurate when an increased number of classi ed hand grasps are included.…”

Wearable upper limb robotics for pervasive health: a review

“…SMG has been studied to estimate and predict human motion intention and motion pattern recognition recently [95,96]. Engdahl et al [97] classified user performance during clinical tests of upper limb transradial procedure, based on analogous SMG spatial features, while exploring the repeatability isolation of SMG control signal over a short period of time during pre-prosthetic training. An experimental comparison was also performed in [98] to evaluate the effect of SMG-based and sEMG-based human-machine interface (HMI) on finger motion classification for more precise control and manipulation.…”