BackgroundHand synergies have been extensively studied over the last few decades. Objectives of such research are numerous. In neuroscience, the aim is to improve the understanding of motor control and its ability to reduce the control dimensionality. In applied research fields like robotics the aim is to build biomimetic hand structures, or in prosthetics to design more performant underactuated replacement hands. Nevertheless, most of the synergy schemes identified to this day have been obtained from grasping experiments performed with one single (generally dominant) hand to objects placed in a given position and orientation in space. Aiming at identifying more generic synergies, we conducted similar experiments on postural synergy identification during bimanual manipulation of various objects in order to avoid the factors due to the extrinsic spatial position of the objects.MethodsTen healthy naive subjects were asked to perform a selected “grasp-give-receive” task with both hands using 9 objects. Subjects were wearing Cyberglove Ⓒ on both hands, allowing a measurement of the joint posture (15 degrees of freedom) of each hand. Postural synergies were then evaluated through Principal Component Analysis (PCA). Matches between the identified Principal Components and the human hand joints were analyzed thanks to the correlation matrix. Finally, statistical analysis was performed on the data in order to evaluate the effect of some specific variables on the hand synergies: object shape, hand side (i.e., laterality) and role (giving or receiving hand).ResultsResults on PCs are consistent with previous literature showing that a few principal components might be sufficient to describe a large variety of different grasps. Nevertheless some simple and strong correlations between PCs and clearly identified sets of hand joints were obtained in this study. In addition, these groupings of DoF corresponds to well-defined anatomo-functional finger joints according to muscle groups. Moreover, despite our protocol encouraging symmetric grasping, some right-left side differences were observed.ConclusionThe set of identified synergies presented here should be more representative of hand synergies in general since they are based on both hands motion. Preliminary results, that should be deepened, also highlight the influence of hand dominance and side. Thanks to their strong correlation with anatomo-functional joints, these synergies could therefore be used to design underactuated robotics hands.