The Madras Diabetes Research Foundation (MDRF) will use artificial intelligence, data science, and machine learning to improve treatment modalities for diabetes. The Foundation has tied up with a technology firm, embedUR Systems, to conduct research. MDRF Chairman V. Mohan said on Thursday that the institution would rely on the data collected from those using continuous glucose monitoring (CGM) patches that could provide more insights as they track glucose variations through the day.
One of the aims is to find if blood glucose fluctuations can help predict type 1 or type 2 diabetes. The data can provide information on predicting complications as well, Dr. Mohan said. MDRF will share the dataset of its readings from the CGM systems given to its patients with embedUR. “Currently, we look at only one parameter, which is blood sugar level. The data will be used to look for patterns that could help doctors improve treatment protocols and predict the complications that may occur,” Dr. Mohan said.
Rajesh C. Subramaniam, founder and chief executive officer of embedUR systems, said CGM was a medical wearable device that used an internet of things framework. From the datasets, a set of markers such as low sugar or high sugar events will be collected and analysed for patterns. The engineers in the company will study the data and develop a model that will be used to predict the health issues that a person with diabetes have specific markers could suffer, he said.
Dr. Mohan drew parallels to the human genome research and later the use of non-hypothesis based approach in genetics to study diabetes, which helped identify key genes and pathways, which had been missed earlier.
