Grant: Doctoral: Vanier Canada Graduate Scholarships

Alberta has one of the highest rates of inflammatory bowel disease, or IBD, in the world. Yet it has received too little attention in our healthcare system. Unfortunately, care is typically provided reactively through emergency departments and hospitals. Diagnosing and providing treatment earlier can significantly improve an IBD patient's health outcomes and quality of life, while reducing costs to the healthcare system. How can we identify and diagnose patients earlier? One promising approach is a statistical technique called machine learning, or ML, which has been applied to healthcare data to detect tumors, diagnose eye and skin diseases. In IBD however, it is largely unexplored. One major limitation in successful ML research is attaining high quality data that contains huge numbers of patients. However, Alberta has rich province-wide clinical databases that follow a population of over 30,000 patients living with IBD. We aim to apply ML techniques to data from Alberta's provincial databases in order to classify IBD patients according to risk and to automatically identify features of the disease on diagnostic images and scans. We will collect demographic, clinical, laboratory, and imaging data. Then we will apply and test our algorithms using best practices. This research could demonstrate a new way to augment the precision of IBD care given by doctors and imaging experts. We hope the long-term result is the deployment of new systems to improve speed and accuracy of diagnostic testing and treatment, and ultimately, patient outcomes. Such systems could also be used as 'second observers', invaluable for trainees, and providers with suboptimal access to experts (such as in rural Alberta or Northern Canada).