Mechanomyogram for identifying muscle activity and fatigue

Facial Muscle Activity Recognition with Reconfigurable Differential Stethoscope-Microphones

Many human activities and states are related to the facial muscles’ actions: from the expression of emotions, stress, and non-verbal communication through health-related actions, such as coughing and sneezing to nutrition and drinking. In this work, we describe, in detail, the design and evaluation of a wearable system for facial muscle activity monitoring based on a re-configurable differential array of stethoscope-microphones. In our system, six stethoscopes are placed at locations that could easily be integrated into the frame of smart glasses. The paper describes the detailed hardware design and selection and adaptation of appropriate signal processing and machine learning methods. For the evaluation, we asked eight participants to imitate a set of facial actions, such as expressions of happiness, anger, surprise, sadness, upset, and disgust, and gestures, like kissing, winkling, sticking the tongue out, and taking a pill. An evaluation of a complete data set of 2640 events with 66% training and a 33% testing rate has been performed. Although we encountered high variability of the volunteers’ expressions, our approach shows a recall = 55%, precision = 56%, and f1-score of 54% for the user-independent scenario(9% chance-level). On a user-dependent basis, our worst result has an f1-score = 60% and best result with f1-score = 89%. Having a recall ≥60% for expressions like happiness, anger, kissing, sticking the tongue out, and neutral(Null-class).

Using a high spatial resolution tactile sensor for intention detection

Intention detection is the interpretation of biological signals with the aim of automatically, reliably and naturally understanding what a human subject desires to do. Although intention detection is not restricted to disabled people, such methods can be crucial in improving a patient's life, e.g., aiding control of a robotic wheelchair or of a self-powered prosthesis. Traditionally, intention detection is done using, e.g., gaze tracking, surface electromyography and electroencephalography. In this paper we present exciting initial results of an experiment aimed at intention detection using a high-spatial-resolution, high-dynamic-range tactile sensor. The tactile image of the ventral side of the forearm of 9 able-bodied participants was recorded during a variable-force task stimulated at the fingertip. Both the forces at the fingertip and at the forearm were synchronously recorded. We show that a standard dimensionality reduction technique (Principal Component Analysis) plus a Support Vector Machine attain almost perfect detection accuracy of the direction and the intensity of the intended force. This paves the way for high spatial resolution tactile sensors to be used as a means for intention detection.

Man-machine interface system for neuromuscular training and evaluation based on EMG and MMG signals

This paper presents the UVa-NTS (University of Valladolid Neuromuscular Training System), a multifunction and portable Neuromuscular Training System. The UVa-NTS is designed to analyze the voluntary control of severe neuromotor handicapped patients, their interactive response, and their adaptation to neuromuscular interface systems, such as neural prostheses or domotic applications. Thus, it is an excellent tool to evaluate the residual muscle capabilities in the handicapped. The UVa-NTS is composed of a custom signal conditioning front-end and a computer. The front-end electronics is described thoroughly as well as the overall features of the custom software implementation. The software system is composed of a set of graphical training tools and a processing core. The UVa-NTS works with two classes of neuromuscular signals: the classic myoelectric signals (MES) and, as a novelty, the myomechanic signals (MMS). In order to evaluate the performance of the processing core, a complete analysis has been done to classify its efficiency and to check that it fulfils with the real-time constraints. Tests were performed both with healthy and selected impaired subjects. The adaptation was achieved rapidly, applying a predefined protocol for the UVa-NTS set of training tools. Fine voluntary control was demonstrated to be reached with the myoelectric signals. And the UVa-NTS demonstrated to provide a satisfactory voluntary control when applying the myomechanic signals.