Machine Methods of Tremorogram Analysis for Human Functional State Estimation

Methods based on the assessment of motor activity by the parameters of functional or pathological tremor are used to determine the functional state of a person. The result of tremor measurements is presented in the form of tremorograms. The result of tremor measurements is presented in the form of tremorograms, they are time series obtained on the basis of tremor amplitudes. Known methods of analyzing nonstationary series can be applied to classify tremorograms by adapting them according to the peculiarities of the initial signals. In this article we propose an algorithm for analyzing tremorograms based on their spectral characteristics and involving spectrum calculation, its smoothing, and comparison of the spectrum with a template based on the DTW method. A variant of Parkinson’s disease diagnostics is shown as an example of application of the proposed algorithm.