Improving Reliability of Myocontrol Using Formal Verification

Abstract

In the context of assistive robotics, myocontrol is one of the so-far unsolved problems of upper-limb prosthetics. It consists of swiftly, naturally and reliably converting biosignals, non-invasively gathered from an upper-limb disabled subject, into control commands for an appropriate self-powered prosthetic device. Despite decades of research, traditional surface electromyography cannot yet detect the subject's intent to an acceptable degree of reliability, that is, enforce an action exactly when the subject wants it to be enforced.In this work we tackle one such kind of mismatch between the subject's intent and the response by the myocontrol system, and show that Formal Verification can indeed be used to mitigate it. Eighteen intact subjects were engaged in two Target Achievement Control tests in which a standard myocontrol system was compared with two "repaired" ones, one based on a non-formal technique, and thus enforcing no guarantee of safety, and the other using the Satisfiability Modulo Theories (SMT) technology to rigorously enforce the desired property. The experimental results indicate that both repaired systems exhibit better reliability than the non-repaired one. The SMT-based system causes only a modest increase in the required computational resources with respect to the non-formal technique; as opposed to this, the non-formal technique can be easily implemented in existing myocontrol systems, potentially increasing their reliability.

Responsible and Regulatory Conform Machine Learning for Medicine: A Survey of Technical Challenges and Solutions

“…The causal ML field is still in its infancy, and many technical challenges remain to be solved [197,198], but this appears to be a very promising direction for future research. 31 Domain knowledge can also be exploited after the fact by formally verifying whether a trained model satisfies application-specific constraints [204,205] or automatically generating safetycritical corner cases for testing the model [206].…”

“…Katz et al [204] employed their proposed techniques to verify application-specific desirable properties of a real-world airborne collision avoidance system for unmanned aircraft. In the medical domain, Guidotti et al [205] have employed an SMT-based technique for formally verifying a desirable input-output property of a prosthesis myocontroller. Because such SMT-based techniques are computationally very demanding and not applicable to arbitrarily large networks, Pei et al [206] have proposed an alternative method (VeriVis) using input space reduction techniques.…”

“…Of course, similar formal verification techniques have also been proposed for other classes of machine learning models, including, in particular, various kinds of tree ensembles [207,208,209]. Several of these formal verification methods can also be used to iteratively "repair" a model that does not (yet) satisfy the specified constraints until it does [205]. Finally, there is a large body of work regarding the formal verification of (e.g.…”

Hybrid Impedance-Admittance Control for Upper Limb Exoskeleton Using Electromyography

“…The advantage of sEMG lies mainly in the fact that it is non-invasive. Examples of the use of sEMG are found in studies that monitor signals to control exoskeletons for arms [4,8,11,16,17,19], hands [9], wheelchair [15], and knee [29]. In [3], the state-of-the-art on lower limb exoskeletons is analyzed.…”

“…Several works combine EMG with other techniques to control exoskeletons, including neuro-fuzzy [8,15], classifiers and neural networks [16], proportional closed-loop control [4], machine learning [17], cinematics [18], and Kalman filter and Hill model [19]. The use of classifiers enables identifying movements with accuracy above 90% [19].…”