Estimating Speed-Accuracy Trade-offs to Evaluate and Understand Closed-Loop Prosthesis Interfaces

Abstract

Objective: Closed loop prosthesis interfaces, combining electromyography (EMG) based control with non invasive supplementary feedback, represent a promising direction to develop the next generation of user prosthesis interfaces. However, we still lack an understanding of how users make use of these interfaces, and how to evaluate competing interfaces. In this study we use the framework of speed accuracy trade off functions (SAF) to understand, evaluate and compare the performance afforded by two closed loop user prosthesis interfaces. Approach: Ten able bodied participants and one amputee performed a force matching task in a functional box and blocks setup at 3 different speeds. All participants were subject to both interfaces in a crossover fashion with a one week washout period. Importantly, both interfaces used (identical) direct proportional control but differed in the feedback provided to the participant: EMG feedback vs force feedback. We thereby estimated the SAFs afforded by the two interfaces, and additionally sought to understand how participants planned and executed the task in the various conditions. Main results: We found that execution speed significantly influenced the performance, and that EMG feedback afforded better performance overall. Notably, we found that there was a difference in SAF between the two interfaces, with EMG feedback enabling participants to attain higher accuracies faster than Force feedback. Further, both interfaces enabled participants to develop flexible control policies, while EMG feedback also afforded participants to generate smoother more repeatable EMG commands. Significance: Overall, the results indicate that closed loop prosthesis interfaces afford users to exhibit a wide range of performance, which is affected both by the interface and the execution speed. Thereby, we argue that it is important to consider the speed accuracy trade offs to rigorously evaluate and compare (closed loop) user prosthesis interfaces.