I am an imaging scientist, data scientist and epidemiologist. Using this unique combination of skills, I study bones, muscles, joints, and fat in the aging population with particular focus on osteoarthritis, osteoporosis, sarcopenia and the interaction of these diseases with one another and with other comorbidities.
Methods and Approaches
My team is focused on developing image segmentation tools (including automated methods in Python, neural networks) to derive metrics that we could apply to standard of care images to improve disease characterization.
We also apply prognostic models, including machine learning data-driven variable selection and validation, to improve our ability to predict clinical outcomes. We have further built expertise in understanding the mechanisms of exposure-outcome associations using causal mediation and structural equation modeling.
Tools I currently use: Python, SAS, R
Projects and Themes
Our team is currently examining the following research aims:
- Understanding the disease process for specific knee osteoarthritis phenotypes in individuals with: a) altered bone quality or bone turnover; and b) with altered metabolism
- Determining imaging-based predictors of success in knee arthroscopic surgeries
- Innovating pixel-level high-dimensional image analysis techniques to discover outcome-sensitive measures
- Elucidating the impact of fat infiltration into muscles and bones on falls, fractures, and frailty
- Evaluating pain and suffering in musculoskeletal diseases while accounting for sex, gender and psychosocial factors
I am currently seeking masters and PhD students with an affinity for data or image analysis, a quantitative background, or interest in machine learning, neural network applications. Those with programming background are highly encouraged to inquire.
Two unique Joint PhD positions (Conferred by U of T and University of Melbourne) are also available. For more information, please see: