Format

Send to

Choose Destination
Neural Netw. 2003 Oct;16(8):1141-60.

A neural network simulating human reach-grasp coordination by continuous updating of vector positioning commands.

Author information

1
Department of Cognitive and Neural Systems, Boston University, 677 Beacon Street, Boston, MA 02215, USA. antonio.ulloa@nih.gov

Abstract

We developed a neural network model to simulate temporal coordination of human reaching and grasping under variable initial grip apertures and perturbations of object size and object location/orientation. The proposed model computes reach-grasp trajectories by continuously updating vector positioning commands. The model hypotheses are (1) hand/wrist transport, grip aperture, and hand orientation control modules are coupled by a gating signal that fosters synchronous completion of the three sub-goals. (2) Coupling from transport and orientation velocities to aperture control causes maximum grip apertures that scale with these velocities and exceed object size. (3) Part of the aperture trajectory is attributable to an aperture-reducing passive biomechanical effect that is stronger for larger apertures. (4) Discrepancies between internal representations of targets partially inhibit the gating signal, leading to movement time increases that compensate for perturbations. Simulations of the model replicate key features of human reach-grasp kinematics observed under three experimental protocols. Our results indicate that no precomputation of component movement times is necessary for online temporal coordination of the components of reaching and grasping.

PMID:
13678619
DOI:
10.1016/S0893-6080(03)00079-0
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center