Overview of Research Interests in Biomedical Engineering

My biomedical research program is primarily focused on biosignal processing, though I participate in some clinical engineering projects involving wireless technologies and have a newly defined interest in Health Information Systems.

In biosignal processing, I investigate signal processing techniques which facilitate the extraction of information from EMG (electrical signals from muscles) and EEG (electrical signals from the brain). I apply artificial neural networks, adaptive filters, time-frequency analysis and machine learning algorithms to measure EMG reliably and extract parameters such as signal conduction velocity and mean frequency. My most recent research uses machine learning to automate biosignal quality analysis during measurement. The goal is to use EMG parameters as diagnostic indicators for neuromuscular pathologies, or as assistive-device inputs.

For a complete list of my publications, see Research Gate or Google Scholar

Projects of Interest

Quality Assessment of EMG - MScE Graduate, G Phillips

Using surface EMG measurements as inputs to assistive devices or as diagnostic indicators requires that the measurements made are reliable. Our work in this area applies machine learning techniques to identify EMG records suspected of being contaminated with noise. This work is part of a collaborative project called cleanEMG which we share with researchers at Carelton University.

• G Phillips, DT MacIsaac, "Pairwise Attribute Noise Detection Applied to Surface EMG," XX Congress of the International Society of Electrophysiology and Kinesiology, Rome , 2014.

 

Muscle Fatigue Assesment - PhD Graduate, Dan Rogers

Many EMG parameters have been shown to track fatigue (conduction velocity, amplitude, mean frequency). Normally though, these parameters can only be measured precisely (in a relative sense) under contraction conditions which maintain constant force and joint angle. Our work in this area applies neural networks, time-frequency analyis, and projection techniques to track fatigue with more sophisticated parameter analysis which facilitates parameter estimation under unrestricted contraction conditions.

• DR Rogers, DT MacIsaac, "A Comparison of EMG-based Muscle Fatigue Assessments during Dynamic Contractions," Journal of Electromyography and Kinesiology, 23(5), pp 1004-1011, 2013.

EMG Measurement Repeatability - MScE Graduate, S Zamman

Using surface EMG measurements as inputs to assistive devices or as diagnostic indicators requires that the measurements made are reliable, and that they are repeatbale. This work investigates the inherent trial-to-trail variability in surface EMG parameter estimation to produce a standard error of measurement benchmark for common EMG parameters such as mean frequency, and for some novel multi-feature EMG fatigue indices.

• SA Zaman, DT MacIsaac, PA Parker, "Repeatability of MES-based Fatigue Assessment in Static and Cyclic Contractions," Proceedings of the 35th Canadian and Biological Engineering Conference (CMBEC'35), Halifax, 2012.