Development of a Three-Dimensional Anthropometric Model for Simulating Hand Work

摘要

This dissertation aims to increasing the understanding on the variables needed to realistically model a human hand by 1) developing a skeleton-driven 3-D parametric model of the hand, and 2) providing new data for predicting hand strength capabilities. Hand models can be used to understand the effect of hand posture and shape to force exertions while interacting with surfaces. 3D models representing skin and bone surfaces and their axes for use in simulation models were developed from Hand-CT Scans. These models were used to adapt four methods for determining phalange Centers of Rotation (CoR). Sphere- and Ellipsoid-fitted CoRs were considered fixed, with pin-joint rotational axes and simple links. 3D-Reuleaux and ICP-based CoRs were considered instantaneous, with variable link lengths affected by the rotation and gliding of the adjacent anatomical segments. There was a significant difference between fixed and instantaneous CoRs, leading to a more accurate and robust kinematic model. These CoRs were leveraged to develop landmark-free statistical models from 43 clinical hand-CT scans. The 3D-surfaces of flat hand posture, previously developed, were used to standardize the data. Kinematics developed from the ICP method were used to rotate finger segments of fitted hands to 1) obtain hand skin measurements in a common posture, 2) predict whole hand skeleton shape/size, and 3) evaluate final shape predictions in their original postures. Principal component analysis and regression (PCAR) were used to develop statistical models for shape/size prediction of individual finger bone geometries, as well as a whole hand skeleton model with hand skin surface reference points for scaling based on anthropometric data (hand length, hand breadth, hand thickness and sex). The predicted skeleton with the skin reference points can be used as a baseline for any hand surface model to establish kinematics based on internal bone segments. Additionally, to complement data in literature, a study of 12 participants was performed to investigate the effect of hand posture and surface orientation on hand force while pressing a flat surface. Joint moment and finger force distribution data from this study can be incorporated in computerized 3D-models of the hand to compare strength capabilities between postures.