Hand rehabilitation is an extremely complex and critical process in the medical rehabilitation field. This is mainly due to the high articulation of the hand functionality. Recent research has focused on employing new technologies, such as robotics and system control, in order to improve the precision and efficiency of the standard clinical methods used in hand rehabilitation. However, the designs of these devices were either oriented toward a particular hand injury or heavily dependent on subjective assessment techniques to evaluate the progress. These limitations reduce the efficiency of the hand rehabilitation devices by providing less effective results for restoring the lost functionalities of the dysfunctional hands. In this project, a novel technological solution and efficient hand assessment system is produced that can objectively measure the restoration outcome and, dynamically, evaluate its performance. The proposed system uses a data glove sensorial device to measure the multiple ranges of motion for the hand joints, and a Virtual Reality system to return an illustrative and safe visual assistance environment that can self-adjust with the subject's performance. The system application implements an original finger performance measurement method for analysing the various hand functionalities. This is achieved by extracting the multiple features of the hand digits' motions; such as speed, consistency of finger movements and stability during the hold positions. Furthermore, an advanced data glove calibration method was developed and implemented in order to accurately manipulate the virtual hand model and calculate the hand kinematic movements in compliance with the biomechanical structure of the hand. The experimental studies were performed on a controlled group of 10 healthy subjects (25 to 42 years age). The results showed intra-subject reliability between the trials (average of crosscorrelation ρ = 0.7), inter-subject repeatability across the subject's performance (p > 0.01 for the session with real objects and with few departures in some of the virtual reality sessions). In addition, the finger performance values were found to be very efficient in detecting the multiple elements of the fingers' performance including the load effect on the forearm. Moreover, the electromyography measurements, in the virtual reality sessions, showed high sensitivity in detecting the tremor effect (the mean power frequency difference on the right Vextensor digitorum muscle is 176 Hz). Also, the finger performance values for the virtual reality sessions have the same average distance as the real life sessions (RSQ =0.07). The system, besides offering an efficient and quantitative evaluation of hand performance, it was proven compatible with different hand rehabilitation techniques where it can outline the primarily affected parts in the hand dysfunction. It also can be easily adjusted to comply with the subject's specifications and clinical hand assessment procedures to autonomously detect the classification task events and analyse them with high reliability. The developed system is also adaptable with different disciplines' involvements, other than the hand rehabilitation, such as ergonomic studies, hand robot control, brain-computer interface and various fields involving hand control.
|Date of Award||20 Oct 2016|
- University Of Strathclyde
|Supervisor||Heba Lakany (Supervisor) & Bernard A Conway (Supervisor)|