A new methodology to improve myoelectric signal processing using handwriting

In this paper, a new biomedical application of handwriting is proposed to enhance the processing of surface electromyographic (SEMG) signals by using the Delta-Lognormal model of the Kinematic Theory of rapid human movements. Three applications are developed from this model: the first two exploit specific time indexes of the end-effector velocity profiles, as recorded by a digitizer. Theses indexes are used 1) to circumscribe the portions of SEMG corresponding to the duration of the rapid movement and 2) to enhance the shape of the SEMG envelops, constructed from a sum of trials collected during repetitive motor tasks. The third application, which does not exploit any kinematic information of the end-effector, aims at decoding the time plan of the muscle activities. Such a plan is assumed to be used in the motor control of smooth complex movements. The resulting tool provides new insights that could be exploited, for example, in the recognition from SEMG envelops of specific patterns from which artificial commands could be constructed to control prosthesis or exoskeleton movements