Abstract
INTRODUCTION: Visually guided reach-to-grasp actions have been proposed to consist of distinct transport and grip components, which may result from the different visual properties required to guide them (location and object size and shape, respectively). However, kinematic studies of hand actions have typically investigated the effect of different conditions on dependent measures one at a time. We examined the interrelationships between various kinematic variables using a multivariate approach. In addition, we developed new measures of grip accuracy and tested how these related to standard reach-to-grasp variables. METHODS: Participants (n=24) performed reach-to-grasp or reach-to-touch movements upon different-sized rectangular objects. Our analysis involved correlating standard kinematic variables (e.g., maximum grip aperture, MGA, and peak velocity of reach, PV, and time at which they occurred, tMGA and tPV) and new grip accuracy variables (shift and orientation of initial grip) with one another across all trials for each subject. We applied a multivariate analysis [using similar logic as representational similarity analysis (RSA) used in fMRI] to test various models of reach-to-grasp components, including a transport-grip model which predicts strong correlations within (but not between) traditional transport and grip and a timing-based model which predicts strong correlations between variables associated with the timing of movement. RESULTS: Our timing-based model of reach-to-grasp movements, but not the transport-grip model, showed strong correlation with the data. Furthermore, as opposed to our hypothesis, participants did not display larger grips on trials where their grip accuracy was low. CONCLUSION: Though standard kinematic measures used in studying reach-to-grasp movements (e.g., PV, MGA) have been theoretically divided into transport and grip variables, these variables do not explain the relationships between variables. The development of new measures of grip accuracy and new applications of multivariate analyses open up new questions for kinematic studies of reach-to-grasp movements in clinical populations and healthy controls.
Meeting abstract presented at VSS 2017