PROJECT: Clinical Knowledge Dissemination

Machine Learning to Provide Intelligent and Customizable Digital Health to Assist Physicians in Underserved Regions

The proposed project aims at providing advanced information intelligence techniques to allow the physicians in rural and underserved regions to enhance their quality of care through readily available intelligent decision systems that utilize mined knowledgebases of medical specialties. A pilot project with collaboration of Division of Gastroenterology and Department of Family Medicine at McMaster University will provide the data for a personalized digital health datacenter at McMaster University. The ultimate goal of this project is to provide data-driven decision support services for rural and underserved regions. Three sources of medical knowledge will be used to create a heterogeneous specialty knowledgebase: i) anonymized specialty patient data will be extracted from electronic health records to supply different data mining and machine learning analyses; ii) research and statistical datasets generated by medical researchers and hospitals provide evidenced-based meta-data to annotate knowledge graph nodes; and iii) expert decisions (diagnosis, treatments, prescriptions) made by medical specialists will provide a rich and authenticated knowledgebase. This specialty knowledgebase is formed as an ontology graph of highly related group of diseases and their shared symptoms. The pattern discovery step utilizes different data mining techniques and machine learning techniques. The intelligent consultant services (ICS) will provide personalized medical information during the patient visit in real-time.

Research Team