Classification characteristics of fine motor experts based on electroencephalographic and force tracking data

Abstract

The application of machine learning techniques provides a data-driven approach for a deeper understanding of the development and expressions of expertise. In extension to the common procedure of comparing experts' and novices' performances in expertise-domain-related tasks we applied conventional classification algorithms. We distinguished between tasks for each participant and between groups, i.e., experts or novices, based on electroencephalographic (EEG) activity patterns and force output variables during four different force modulation tasks. The tasks under investigation involved sinusoidal and steady force tracking tasks, which were performed with the left and right hand. Classification of tasks based on EEG patterns as well as force output was possible with high accuracy in novices and experts, whereas classification of group membership, i.e., experts or novices, was at chance level. In follow-up analyses, we found a high degree of individuality in the EEG patterns of the experts, implying the long-term development of specialized central processing during fine motor tasks in fine motor experts. Taken together, the results suggest that continuous practice in the work context leads to the development of a highly individual and task-specific central control pattern.