CASPIAN JOURNAL

MANAGEMENT AND HIGH TECHNOLOGIES

Mathematical processing of tenzotremorogramms: methods and models

Read

Zhvalevsky O. V. Mathematical processing of tenzotremorogramms: methods and models // Caspian journal : management and high technologies. — 2018. — №2. — pp. 149-161.

Zhvalevsky O. V. - Research Associate, St. Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences, 39 14-th Line, VI, St. Petersburg, 199178, Russian Federation, ozh@spiiras.ru

The paper deals with Parkinson's disease diagnostics automation, based on mathematical analysis of physiological signals. The aim of automation is to develop equipment for objective, accurate and, if possible, early Parkinson's disease diagnostics. The study shows that each of these tasks of diagnostics requires different equipment. For objective diagnostics there can be used a simple mobile application available for any smartphone with a set of standard sensors. For accuracy of diagnostics, specialized hardware and software are used. Specialized hardware allows to register high quality experimental data, and to organize full measurement plan, and specialized software implements effective methods of machine learning and mathematical simulation. Early diagnostics is possible if based on complex measurements accompanied with non-motor manifestation of Parkinson's disease. The study shows the effectivity of developing Parkinson's disease diagnostics automation system based on mathematical analysis of tensotremorogramms. Tensotremorogramm is the result of tremor registration using a special procedure developed by S.P. Romanov for objective evaluation of state of movement construction system (N.A. Bernstein). The main object of investigation is to develop a recognition system based on mathematical analysis of tensotremorogramms. Therefore, a multilayer schema of system recognition building is proposed; the place of machine learning and mathematical simulation in this schema is also shown.

Key words: болезнь Паркинсона, диагностика заболеваний, автоматизация, функциональное состояние, тензотреморограмма, анализ временных рядов, машинное обучение, математическое моделирование, система распознавания, концептуальная модель, Parkinson's disease, disease d