To be able to extract meaningful information from physiological data quickly and accurately, automated and sophisticated artifact identification and removal tools are required. We developed a graphical user interface (GUI) application for Matlab, developed for inspection and identification of artifacts in physiological recordings. The tool allows concurrent visualisation of two parallel channels, the fist identifies artifacts within the physiological data and the second is used to identify points of interest from Electroencephalography (EEG), Electrocardiography (ECG) – R peaks, Respiration, etc. During the analysis, the user may choose to identify and mark the artifact regions which (although present) would be ignored from the further data analysis. The pre-processed signals are saved and displayed as a whole, which can be easily handled for further analysis using GUI controls. The GUI provides a starting point for artifact removal analysis for any physiological signals and helps in speed up the data-cleaning process considerably.
The number of upper-limb amputees in Australia is currently around 35,000, or 0.15%, of the population. This is a significant rise from 0.1% in 1993 (Australian Bureau of Statistics, 1993), and is due to rising levels of diabetes and cardiovascular disorders. Existing prosthetic hands are usually electrically powered and are controlled by the myoelectric recording from the surface of the residual muscles of the amputated end. However, the obstacle in the realisation of the true potential of these devices is that the myoelectric controller is unable to identify individual finger control commands. This project aims to investigate how prosthetic hand devices can provide the user with the dexterity of using individual fingers separately and to develop an intelligent myoelectric prosthetic hand control system using only a few sensors.
Figure 1: The overall schematic of proposed system
In depth understanding of the underlying muscle activities and neural mechanisms are potentially vital for reversing the effects of stroke and for better design of prostheses. By identifying the synergy activations that are affected following a stroke or amputation, it is possible to develop focused rehabilitation methods that specifically train the impaired synergies.
Hence, the objective of this project is to identify important neural synergies that lead to the development of novel rehabilitation methods for stroke survivors and amputees during reaching and grasping tasks. As a chief investigator (leader) of this project, I have already developed novel ICA and Nonnegative Matrix Factorisation (NMF) based source separation and muscle synergy algorithms and identified important synergies for grasping and reaching tasks, this would help me to identify the weak neural synergies and develop targeted stroke rehabilitation technique.