Jim Reilly works at the interface of machine learning and signal processing applied to health related problems, particularly in neuroscience and psychiatry. Particular thrusts are development of improved machine learning algorithms, diagnosis and treatment of psychiatric illness, prognosis for coma outcome, and assessment of infant motor movement relating to neurological deficit.


Narges Armanfard

Dr. Reilly

Dr. Narges Armanfard obtained her PhD degree from McMaster University, Canada in 2016. The focus of her PhD research was developing effective machine learning algorithms for data classification. In this vein, she proposed and developed the novel concept of localized feature selection and classification that its performance is demonstrated for classification in complex and high dimensional sample spaces. She also developed a novel machine learning based system for automatic and continuous coma patient assessment and outcome prediction.

Her current research projects with Dr. Alex Mihailidis involve the design and application of machine learning techniques for autonomous and unobtrusive home-based physiological signal monitoring of elderly adults using the information recorded by ambient sensors.

Phil Chrapka

Dr. Reilly

Rober Boshra

Dr. Reilly

Rober's primary research topic pertains to local brain potentials, recorded from the human cortex (electrocorticography), and their relation to both scalp event-related potentials and high frequency brain oscillations. Rober is interested in the analysis of such recorded signals in conjunction with state-of-the-art machine learning and digital signal processing techniques (primarily machine learning) to elucidate facets of consciousness, cognition, and language processing.

Omar Nassif

Omar Nassif is a Master's student co-supervised by Prof. Jim Reilly in the Dept. of Elec. and Comp Eng, and Prof. Vickie Galea, School of Rehabilitation Science. He is working on characterization of premature infant motor movement for assessment of potential neurological deficit. The proposed methodology is to extract movement primitives from video sequences for analysis by machine learning methods.