Design and implementation of a physiological wrist simulator with applications to surgical reconstructions

Abstract

The wrist is involved in almost one-third of all traumatic injuries, which lead to pain, joint instability and loss of functionality. A physiological wrist simulator replicates the kinematic and kinetic conditions of the human wrist in cadaveric specimens, thus enabling the quantification of the biomechanics of a healthy wrist, as well as allowing a controlled comparison of the biomechanical effects of wrist pathologies and surgical reconstructions in vitro. Owing to the presence of muscle redundancy and co-activation in a natural joint, it is vital to optimize the load distribution between the various muscles in a simulator, in order to replicate physiological wrist motions. Typically, control strategies in the literature employ one of either position or force feedback to solve this indeterminate problem; however, it was hypothesized that a combination of these two types of feedback would result in a more physiologically realistic output. To test and iteratively improve the various control strategies, a phantom – an artificial, functional replica of a human forearm and hand – was created. Movement of the wrist was driven by electromechanical actuators connected to the hand via steel cables replicating the six primary muscles of the wrist – flexor carpi radialis (FCR), flexor carpi ulnaris (FCU), extensor carpi radialis longus (ECRL), extensor carpi radialis brevis (ECRB), extensor carpi ulnaris (ECU) and abductor pollicis longus (APL). An optical motion capture system was used to record joint angles, while load cells were used to monitor muscle forces. Two novel control strategies combining position and force feedback – hybrid control and cascade control – were developed and tested on the phantom, and were observed to be superior to position or force control alone, based on repeatability, low kinematic error and ability to induce co-contraction. Following the phantom study, planar and non-planar wrist motions were simulated on ten cadaveric specimens using hybrid and cascade control. High repeatability and low kinematic error observed across specimens reflected the robustness of these control strategies. The simulator design enabled the wrist to be tested in various hand orientations – vertically upwards (hand above the elbow), vertically downwards (hand below the elbow) as well as horizontal (palm facing the ground) – the latter of which exhibited higher extensor forces as compared to the two vertical positions (p < 0.017). Interestingly, in the case of circumduction, muscle forces of the flexors were dependent on the direction of movement, with the peak FCR force being higher for clockwise circumduction by 27% (p = 0.017), while that of the FCU being higher for anticlockwise circumduction by 40% (p = 0.013). Once the biomechanics of the normal, intact wrist joint were quantified, pathologies were simulated to observe their effects in altering wrist biomechanics. The FCR and the FCU are regularly used for tendon transfers to treat trauma, arthritis or nerve disorders at the wrist; these were simulated by actuating the specimens without connecting the FCR and FCU, individually. In the absence of FCR or FCU, the mean force of the ECRB decreased by more than 30% (p < 0.05), thus causing a larger drop in extensor force than the compensatory rise in the remaining flexor force, which might possibly indicate loss of muscle strength and joint instability at the wrist. Another study looked at the treatment of osteoarthritis at the base of the thumb, wherein trapeziectomy, i.e. excision of the trapezium, was compared to three types of surgical reconstruction performed to fill the void of the excised bone – suture suspension arthroplasty, tendon tie-in implant and intermetacarpal ligament reconstruction. A 112% rise in the APL force during flexion-extension in trapeziectomy (p = 0.011) suggested that leaving the trapezial gap unfilled could lead to muscle fatigue and joint instability. However, it was observed that the prosthetic implant and the ligament reconstruction successfully restored wrist biomechanics, which could lead to improved clinical outcomes. Thus, a robust physiological wrist simulator was designed, tested and used to investigate the effects of wrist pathology and surgical reconstructions on musculoskeletal biomechanics.