Properties of window dispersion increments of a myogram as a stochastic process

One of the most rapidly progressing directions in the modern medicine is brain activity research. There is a set of techniques of carrying out experiments in this area. One of the approaches is the so-called method of caused potentials: a testee repeats some actions and his brain activity and some auxiliary signals are registered for further analysis. The authors of this paper and their colleagues published several papers about algorithms of reference points localization by a myogram (and some other signals) for the experiment with finger movements. Accuracy of detection of these points is critical for processing a magnetoencephalogram — signal average by multiple epochs centered by them is computed to find a sensor with the best response (it should be closer to the source of activity than other). In this paper, some theoretical aspects of these algorithms such as a myogram and increment of its window dispersion as random processes are discussed.