A framework for quantifying the mechanics of dexterous grasp
Sobinov, A. R.; Ma, X.; Okorokova, E. V.; Greenspon, C. M.; Raman, C.; Shelchkova, N.; He, Q.; Darabi, N.; Bhatt, R.; Jordan, P.; Arters, P.; Hatsopoulos, N. G.; Miller, L. E.; Bensmaia, S. J.
Show abstract
AO_SCPLOWBSTRACTC_SCPLOWA hallmark of primate behavior is the exceptional ability to dexterously grasp and manipulate objects, yet the investigation of the neural mechanisms that support manual dexterity has been hindered by technical challenges. Optical hand tracking is complicated by frequent occlusions, and contact forces are hard to measure with sufficient precision. Furthermore, while monitoring the kinematics during reaching phase, and the contact forces during object manipulation phase is difficult, joint torques are impossible to measure directly. While challenging, the ability to estimate joint torques in the complex primate hand could provide an invaluable continuous mechanical description spanning both phases. With this in mind, we have developed an experimental apparatus and data processing pipeline for quantifying these variables describing prehension. The apparatus presented objects of various sizes and orientations throughout the workspace, evoking different grasping strategies. Object surfaces were instrumented with thousands of pressure-sensitive elements, enabling high-resolution measurement of distributed contact forces. Simultaneously, eight high-speed cameras were used to reconstruct hand and arm movements with markerless tracking, triangulating 3D landmarks, and mapping them onto a musculoskeletal model, enabling estimation of time-varying joint angles. This posture quantification allowed contact forces to be automatically assigned to specific hand segments, in close agreement with manual human annotations. We used the reconstructed movements and contact forces with the musculoskeletal model of the hand to compute inverse dynamics, yielding joint torques throughout behavior, unifying the hand kinematics and grasp forces into a single physical description. Throughout the processing, we identified individual neurons in the motor cortex of monkeys that were related to grasp force, kinematics, and torques. Together, this framework enabled a comprehensive and precise physical characterization of primate manual behavior, providing a foundation for investigating the neural mechanisms of manual dexterity.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Hand Knob Area of Motor Cortex in People with Tetraplegia Represents the Whole Body in a Modular Way 96%
- Flexible regulation of representations on a drifting manifold enables long-term stable complex neuroprosthetic control 94%
- Task-driven neural network models predict neural dynamics of proprioception 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.