Health Science Seminar Series

This interdisciplinary research seminar is presented by the School of Physical Therapy, the College of Kinesiology, and the Office of the Associate Vice-President Research – Health. 

Thursday, Nov. 23
9-10:30 am
Health Sciences Building GD04

Lecture title: Cluster analysis with self-organizing maps: Subgrouping clinical patients according to movement characteristics

Abstract: Artificial neural networks are a powerful artificial intelligence technique currently employed in the medical field to subcategorize patient populations. Self-organizing maps (SOMs) are a special class of artificial neural networks and provide a low-dimensional visualization of high-dimensional data. An important (non-linear) feature of SOMs is their topology preservation, which means similar map regions represent similar data. This seminar will discuss a study designed to identify and characterize clusters in a heterogeneous low back pain sample of high-dimensional kinematic data collected during a modified Trendelenburg task. Identification of subgroups within the database could help elucidate aberrant movement functionality and facilitate future diagnostic model building. 

Guest speaker: Melanie Bussey, PhD, University of Otago, New Zealand

Professor Bussey is a researcher in clinical biomechanics, and serves as director of the Motion Analysis Laboratory and chair of graduate research in the School of Physical Education Sport and Exercise Sciences at the University of Otago. Recently appointed as an adjunct professor in the School of Physical Therapy at the University of Saskatchewan, her specific research expertise is the relationship between disturbed postural stability and pelvic girdle pain mechanisms, sub-group evaluation and discrimination utilising biomechanical, motor control and clinical tools. Her work has been published widely and funded by the Health Research Council of New Zealand, Otago Medical Research Foundation, Maurice and Phyllis Paykel Trust, Sport Science New Zealand and the Spondyloarthritis Research Consortium of Canada.