Development of a Feedback-Controlled Elbow Simulator: Design Validation and Clinical Application

This work involves three topics that advance the functionality of an elbow simulator in the Orthopaedic Biomechanics Laboratory at Allegheny General Hospital. To draw clinically and scientifically meaningful conclusions from future cadaver studies conducted with the simulator, its design must be validated and the accuracy of the data collection methods demonstrated. The simulator was designed to offer physiologically-correct adjustable moment arms throughout the elbow's range of motion. To validate this, muscle moment arms were measured in three cadaver elbow specimens. Flexion-extension moment arms were measured at three different pronation/supination angles: fully pronated, fully supinated, and neutral. Pronation-supination moment arms for four elbow muscles were measured at three different flexion-extension angles: 30, 60, and 90. The numeric results compared well with those previously reported. The biceps and pronator teres flexion-extension moment arms varied with pronation-supination position, and vice versa. This represents the first use of closed-loop feedback control in an elbow simulator, one of the first reports of both flexion-extension and pronation-supination moment arms in the same specimens, and demonstrates the adjustability of the moment arms that the elbow simulator can produce. Towards accurate motion analysis of the radial head, two areas were investigated. The first identified the phenomena of camera-switching, which occurs in motion analysis when data from one or more cameras is temporarily excluded from the computation of a marker's three-dimensional position. Tests with static markers showed that camera-switching could cause up to 3.7 mm of perceived movement. The second area of investigation set the stage for future studies with cadaver elbows. A protocol was developed to quantify both the travel of the native radial head, radial head implants, and the finite helical axis during pronation-supination movement. The tracking of implant motion employs a unique circle-fitting algorithm to determine the implant's center. A video-based motion analysis system was used to collect marker position coordinates actuated by a precision micrometer table. MATLAB code was designed and implemented to compute both the radial head position and finite helical axis from these data. Immediate future work will use these algorithms to evaluate radial head implants in comparison to the native radial head.