Upper limb joint kinematics resulting from a model conceived for the analysis of SHAP motor tasks performed by trans-radial amputees: discrepancies with the Vicon upper limb model joint kinematics

Introduction Trans-radial amputees often rely on compensatory movements for their daily activities. Numerous studies have explored upper limb kinematics using optoelectronic systems and biomechanical models to evaluate these compensations [1]. This study aims to develop a novel upper limb kinematic model (adapted from [2]) suitable for trans-radial myoelectric prosthesis users. As a preliminary investigation, this model was compared with an existing commercial upper extremity model in a healthy population. The proposed model is not intended to describe forearm prono-supination, which will require further investigation. Methods Eight healthy individuals (3 females; age: 39±11 years) performed three repetitions of four manual tasks extracted from the South Hampton Assessment Procedure (SHAP): lifting a light object, carton pouring, lateral grip, supine extension grip. A 10-camera optoelectronic system (Vicon, UK) was used to reconstruct the proposed unilateral upper extremity 4-segment model using a 20-marker set: on the thorax (4), shoulder (3), arm (6), forearm (5), and hand (2). Arm and forearm segments included 2 calibration markers each. To estimate shoulder joint kinematics, the glenohumeral joint center is determined using a functional method (SCoRE) [3]. Reference systems and kinematics definitions were based on the International Society of Biomechanics guidelines [4]. Comparisons between the upper limb joint kinematics obtained from the proposed biomechanical model and those from the Vicon Upper Limb model [5] were conducted. For both models, Range of Motion (ROM) for each joint degrees of freedom (DOF) was calculated during the reaching phase of each task, and the comparisons were performed using the Bland-Altman analysis. Results The two models showed very good agreement, with the absolute mean difference in ROMs being ≤2° across all DOFs and tasks, and the limits of agreement of up to ~5° (Table 1). The lowest level of agreement was found in the shoulder intra/extra-rotation during the lateral grip task, where limits of agreement ranged between -8° and 7°. Discussion The current findings show a high degree of similarity between the two upper limb models when comparing overall kinematic parameters such as ROM during manual tasks in healthy individuals. In the proposed model, marker calibration is performed to facilitate the application of the kinematic model to trans-radial amputees, addressing the challenges of placing markers on prosthetic parts that resemble anatomical landmarks, especially during dynamic tasks. Future developments will focus on acquiring kinematic data from amputees to refine the model.