Fuzzy c-means clustering analysis of the EMG patterns of six major hand grasps

Abstract

We seek to quantify the electromyographic similarities and differences between six major grasping patterns of the human hand: cylindrical, hook, palmar, lateral, spherical, and tip. Fuzzy c-means clustering (FCM) is a data reduction technique that allows continuous membership, on a [0,1] interval, of data points in the representative clusters. We found no grasp to be completely distinct from any other grasp, and found significant overlap of characteristics between several grasps. This overlap is quantified and discussed.