A multi-rigid-body system model for thorax and upper-limb joint kinematics estimation designed for prosthetic users with trans-radial amputation

Upper-limb joint kinematics has often been estimated using customized approaches tailored for specific research questions. While multiple features are critical for an accurate estimate of upper-limb kinematics, only a few models combine them, and none has been applied to trans-radial prosthetic users. This study aimed to develop a multi-body system model of the thorax and upper-limb joint kinematics, aligned with the International Society of Biomechanics guidelines, applicable to both trans-radial prosthetic users and able-bodied individuals. The model's performance was compared to a commonly used commercial model. One able-bodied individual performed seven 1-Degrees-of-Freedom movements; seven able-bodied individuals and six trans-radial amputees using a myoelectric prosthesis with an articulated wrist performed five unilateral activities of daily living (ADL)-like tasks. Marker positions were recorded with a motion analysis system. Shoulder, elbow and wrist kinematics estimated with the proposed model were qualitatively and quantitatively compared with those from the commercial model. Shoulder kinematics were consistent between models across all tasks. Differences in elbow and wrist kinematics were mainly due to distinct landmark identification and segment definitions. In able-bodied individuals, the proposed model yielded reduced forearm pronation/supination ROMs but more coherent wrist deviation measures compared to the commercial model. For prosthetic users, the proposed model showed larger discrepancies at the shoulder and elbow, and smaller at the wrist, relative to able-bodied estimates. This study introduces a versatile multi-rigid-body system model for joint kinematics estimation tailored for trans-radial prosthetic users and provides a framework for comparing joint kinematics with existing normative and pathological datasets.