A technique for motion capture of the finger using functional joint centres and the effect of calibration range of motion on its accuracy.

A method of kinematic analysis of the fingers using stereo-photogrammetry, referred to as the phalanx transformation technique, has been proposed. Functional methods were used to define the joint axes and subsequently each finger segments' anatomical coordinate system. Thirteen subjects were tested and the accuracy of the technique assessed. The average error across the three joints of the finger was found to be 0.6 mm, which translates to a 2.2% error in predicted joint reaction force when using a biomechanical model. The subjects were required to have sufficient movement in their joints to define the joint axes functionally. Some subjects of clinical interest can have a significantly reduced mobility owing to injury or pathology, therefore, the effect of calibration range of motion on accuracy was analysed. It was found that, for a range of motion typical of a subject with rheumatoid arthritis, the errors in predicted joint reaction force were < 7%. The accuracy of this technique compared favourably with others previously proposed and, considering the other errors inherent in modelling, those found in this study were deemed to be acceptable.