Pilot study on fine finger movement regression, using FMG
Rana Sadeghi Chegani, Carlo Menon · 2017
Predicting hand gestures and finger movements include a wide range of applications in different fields such as human computer interaction, rehabilitation, and prosthesis control. Research in this area, mainly focuses on hand gesture classification which limits the ability of the system to a set of predefined gestures and also limits the ability to control fine finger movements. Force Myography (FMG) is a novel method in which the volumetric change of the muscles associated with a functional motor movement is measured. In this study, the feasibility of using the FMG signals for predicting fine finger movements, and the effect of the hand movement on the prediction was investigated. To obtain the FMG signals, an array of 16 Force Sensing Resistors (FSR) was utilized. To record the trajectory of finger movements, eight calibrated infrared cameras were used. Ten reflective markers, were placed on the index and middle fingers, the thumb and the back of the hand. The FMG signals and the location of the markers were collected while the participant placed their hand in three different predefined locations parallel to the sagittal plane passing through humerus and performed three different hand gestures. FMG signals were collected, and the marker trajectories were fed into a Random Forest Regression algorithm. The results showed an average squared correlation coefficient higher than 75%, on different hand gestures and locations, which proves the feasibility of using FMG signals to predict fine finger movements, in three predefined locations, for three different hand gesture.