Sunil Kalmady Vasu, PhD

Area of Study / Keywords

Medical AI Cardiovascular AI Computational Psychiatry Machine Learning Clinical decision support Biostatistics Electrocardiogram Medical Imaging functional MRI

About

My research integrates machine learning and data science to advance personalized and predictive strategies for tackling complex medical disorders. I hold a Master’s degree in Medical Biotechnology with a specialization in Human Genetics, a Master’s degree in Computer Science, and a Ph.D. in Neuropsychiatry focusing on brain imaging in mental disorders. I completed my postdoctoral training at the Alberta Machine Intelligence Institute (Amii), one of Canada’s leading centers of excellence in AI research.

This interdisciplinary foundation enables me to approach biomedical problems from both biological and algorithmic perspectives. Over the past decade, I have co-supervised graduate students and international research interns across computing and biomedical domains, guiding projects that bridge methodological innovation with clinical application. My team and I have developed, validated, and deployed machine learning models using large-scale structured and unstructured healthcare datasets, including clinical records, imaging, and physiological signals.

I aspire to create intelligent systems that meaningfully impact patient outcomes, transforming diagnosis and personalized treatment. I am equally passionate about mentoring and collaboration, guiding researchers who shape the future of AI-driven medicine.

Research

As an Adjunct Professor of Computing Science and Senior Machine Learning Specialist at the University of Alberta, I lead a research team developing machine learning models that predict diagnostic and prognostic outcomes in cardiovascular disease. Using large-scale datasets of electronic medical records, ECGs, and echocardiograms spanning over one million patient records, our work has appeared in Nature Digital Medicine and other leading journals. Previously, my research in psychiatric disorders used multimodal neuroimaging and machine learning to model symptom clusters and treatment responses in schizophrenia and OCD, with findings published in Nature Schizophrenia and featured in the scientific press.