predict the risk of type 2 diabetes based on early stage and late stage risk factors: genomics, bio-marker, lifestyle, MRI image and environmental data using a deep learning model
Principal Investigator:
Mr Hancheng Zheng
Approved Research ID:
49322
Approval date:
May 14th 2019
Lay summary
The aim of this project is to build an early-stage type 2 diabetes risk model using both early-stage risks factor, such as genetic data, lifestyle, environment data, and late-stage risk factor, such as the bio-marker, family history etc. There has been some basic statistic model to predict risk of diabetes based on classic late-stage risk factors, such as waist circumstances, fasting blood glucose, fasting insulin in addition with family diabetes history. Recently, one artificial intelligence model has improved prediction power by at least 50% by using extra 900 selected risk factors from insurance data, comparing to just using limited classic diabetes risk factors. Therefore, we plan to build a more powerful risk prediction model with artificial intelligence model, which recently has been widely used in biology research and healthcare services, because it is particularly well suited for data enriched complex problem. The period of this project will be one year. The expected value of this result will be to quantify the changes in the risk of type 2 diabetes given a lifestyle change for any individual. Therefore, people can get more precise health advice on how to manage their risk of type 2 diabetes based on people's personal genomic, bio-marker , abdominal MRI imaging data, and lifestyle condition.