Forecasting Levodopa-Induced Dyskinesia in Human Subjects with Parkinson’s Disease

We have developed a novel data augmentation procedure that significantly enhances machine learning-based classification of different brain imaging scans. Having successfully demonstrated proof-of-concept in a rodent model, we are now expanding this approach to clinical applications in humans. Specifically, we aim to utilize this technology to identify early biomarkers of neurodegenerative diseases, enabling personalized treatment strategies based on individual disease progression rates.

As a starting point, we will focus on Parkinson’s disease, which affects approximately 100,000 Canadians and is the second most prevalent neurodegenerative disorder. Over half of patients develop levodopa-induced dyskinesia, a challenging motor side effect. Our previous research demonstrated that individuals who develop this side effect exhibit distinct brain activity patterns from those who do not—even on the first day of levodopa treatment.

With our advanced machine learning technology, we believe we can accurately identify “at-risk” patients before symptoms manifest. This capability will provide clinicians with a powerful tool to tailor treatment strategies, mitigate dyskinesia risk, and uncover novel therapeutic targets. Ultimately, this project aims to establish imaging-based classification as a clinically viable approach for early-stage disease detection and personalized intervention.

Faculty Supervisor:

Ji Hyun Ko

Student:

Partner:

Cubresa Inc

Discipline:

Life Sciences

Sector:

Manufacturing; Professional, scientific and technical services

University:

University of Manitoba

Program: