![]() In many of these studies of both human and nonhuman primate hand movements, principal component analysis (PCA) has been applied to show that most of the correlated rotation of multiple DoFs can be explained by a much smaller number of principal components (PCs). Even when normal subjects are instructed to move one finger, correlated motion occurs in the adjacent fingers ( Hager-Ross and Schieber 2000). 2003), or haptic exploration ( Thakur et al. 1997), producing a sign language alphabet ( Jerde et al. Simultaneous correlated motion at multiple DoFs also is present during more sophisticated uses of the hand, such as typing ( Soechting and Flanders 1997), playing the piano ( Engel et al. 1998, 2002 Santello and Soechting 1998) and in monkeys ( Mason et al. Grasping an object, for example, entails simultaneous motion at multiple joints, with correlated rotation of multiple DoFs, both in humans ( Santello et al. Natural movements of the hand rarely involve motion at a single joint or a single rotational DoF. Skeletal joints thus provide 24 rotational DoFs for the hand. ![]() The wrist allows flexion/extension and abduction/adduction, and the entire hand can be rotated in pronation/supination produced by movement of the radius relative to the ulna. The carpometacarpal joint of the thumb allows flexion/extension, abduction/adduction, and opposition. For example, the metacarpophalangeal (MCP) joints of the fingers allow both flexion/extension and abduction/adduction. Although the proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints have only one degree of freedom (DoF), other joints in the hand rotate about more than one axis. Fifteen joints are present in the 5 digits, 16 if the wrist is considered as well. Neural control of hand movement is complex, in part due to the hand's large number of mechanical degrees of freedom ( Brand and Hollister 1993 Hartman and Straus 1933 Tubiana 1981). We suggest that the motor cortex might act to sculpt the synergies generated by subcortical centers, superimposing an ability to individuate finger movements and adapt the hand to grasp a wide variety of objects. Although PCs, jPCs or isomap dimensions might provide a more parsimonious description of kinematics, our findings indicate that the kinematic synergies identified with these techniques are not represented in motor cortex more strongly than the original joint angles. But the remaining PCs and jPCs were predicted with lower accuracy than individual joint angles. In decoding analyses, where spikes and LFP power in the 100- to 170-Hz band each provided better decoding than other LFP-based signals, the first PC was decoded as well as the best decoded joint angle. For most spike recordings, the maximal absolute cross-correlations of firing rates were somewhat stronger with an individual joint angle than with any principal component (PC), any jPC or any isomap dimension. We used PCA, jPCA and isomap to extract kinematic synergies from 18 joint angles in the fingers and wrist and analyzed the relationships of both single-unit and multiunit spike recordings, as well as local field potentials (LFPs), to these synergies. To examine the possibility that motor cortex might control the hand through such synergies, we collected simultaneous kinematic and neurophysiological data from monkeys performing a reach-to-grasp task. ![]() A few kinematic synergies identified by principal component analysis (PCA) account for most of the variance in the coordinated joint rotations of the fingers and wrist used for a wide variety of hand movements.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |