ALGORITHM FOR DETECTING FINE MOTOR DEFECTS BASED ON INFORMATION FROM SMARTPHONE SENSORS

Abstract

Digitalization is the leading trend of modern humanity. It allows you to solve many everyday tasks with the help of devices with specialized algorithms, facilitating everyday life, as well as solving a number of tasks for which qualified specialists were needed yesterday. One of these tasks is the independent preliminary diagnosis of patients in medicine. The ability to perform such diagnostics allows you to reduce the time to identify problems with various diseases, in particular neurological disorders, including cases such as a defect of fine motor skills, this allows you to reduce the burden on medical specialists. It is worth noting that time plays a crucial role in the process of providing medical care, and the timely provision of medical care can save a person's life. Thus, the development of a solution that allows independent preliminary diagnosis of fine motor defects by using technical tools that almost everyone has is an urgent task today. The aim of the work is to expand the methods for diagnosing the presence of defects in fine motor skills. To achieve this goal, the tasks were set to study the available solutions on the topic and develop a specialized algorithm intended for use in smartphones as part of a biomedical monitoring system. The article presents an algorithm for determining the defects of fine motor skills of a person according to the kinematic sensors of a smartphone – a three-axis accelerometer. The presented solution is based on the analysis of the deviation angles obtained from the smartphone accelerometer when the patient performs the assigned task (exercise). The task requires the patient to take a starting position for three seconds and then hold the smartphone at arm's length for 10 seconds, during which the readings of the three-axis accelerometer are measured. The test results of the solution showed the accuracy of the solution at the level of 0.05 of the alpha error and 0.09 of the beta error. The results obtained indicate the possibility of using the solution for preliminary self-diagnosis and can be used as an element of the diagnostic module in large biomedical monitoring systems.