Conceptual Design of a Multi-Sensor based Dexterity Assessment System

Evaluation of manual dexterity is very important in determining the hand functionality and degree of limitation in activities of daily living. Various tests have been developed for the assessment of manual dexterity. During the application of these tests, there is an expert who determines the duration for each movement while the participant performs standardized tasks. This leads to patients being given appointments days later, especially in overpopulated cities. In addition, various errors may occur due to the human factor in the measurements. In this study, we proposed a system that can calculate the thinking time, disc placement time, disc rotation time and total test time to produce dexterity score during the administration of the Minnesota Manual Dexterity Test (MMDT). The aim of the study is to perform dexterity assessment autonomously with high accuracy using multiple sensors and accordingly reduce the workload of experts and measurement errors. For the autonomous determination of dexterity, data obtained from electromyograph, IMU sensors and a camera were used. The collected data is processed and a score is obtained, and according to this score, an expert system rule base is used to make inferences about the dexterity level of the participant and presented to the experts.