The Biostatistics and Health Data Science lab develops cutting-edge statistical methods for effectively leveraging information  from complex and multi-modal data produced by high-tech medical imaging and wearable device technologies. The overarching clinical goal of the newly developed statistical methods is three-fold:

 i) to  facilitate discovery, evaluation and validation of novel non-invasive biomarkers

i

i) to improve prediction of disease risk and health condition

iii) to delineate complex pathophysiology of various diseases for their cure and prevention.

The specific data modalities our group studies encompass traditional types that are categorical-, ordinal- and continuous-scaled, as well as those that are highly structured and complex, such as high-dimensional, multi-dimensional, functional (e.g., curve and image) data and their mix.

​

Current Projects:​

1. Develop statistical  methods that can effectively analyze color intensity data of patient-sourced fingernail photos to enable non-invasive detection of anemia in infants.

​

2. Develop statistical methods that leverage glucose level data collected by wearable glucometers to dynamically predict impending adverse events (e.g., hypoglycemia) in diabetic patients.

​

3. Use renogram image data to non-invasively detect kidney obstruction.

​