The rapid advances in technology and widespread adoption of Electronic Health Records in healthcare has created an abundance of clinical data, and has faciliated for the development of algorithms to model human health. Taking inspiration from this, I am particularly interested in using deep learning methods to model patient’s trajectory and to predict adverse clinical events in the ICU. I am developing deep learning based models to predict Sepsis - a medical condition characterized by whole-body inflammation which affects between 800,000 and 3.1 million Americans, and Atrial fibrillation - which is often characterized by an irregular heart rate.
Apart from performing retrospective clinial analysis, I care about deploying these algorithms in the real world. And to this end our lab is actively developing real time streaming platforms for deployment of our models in the ICU.
- Predictive analytics in Healthcare
- Applied Deep Learning
- Google Cloud – ML Engine
- Signal Processing
- Multivariate time series
- Data visualization
- ECG, PPG and accelerometer data analysis
- Information & graph theory