Speaker: Cao (Danica) Xiao, PhD, Research Staff Member, AI for Healthcare, IBM Research
Date: Thursday, January 24, 2019
Time: 03:00pm – 04:15pm
Location: Jesse W. Mason Building, Room 2117
Abstract: The massive generation and collection of digitized health data, e.g. electronic health records (EHRs), along with these notable, longstanding healthcare deficiencies including diagnostic errors and medication mistakes are calling for more advanced techniques to power health data for improving healthcare. Deep learning provides great potential in transforming diverse health analytics tasks. In this talk, we mainly explore the following ones:
- Accurate medical embedding across correlated and multimodal medical entities to facilitate disease diagnosis or adverse drug reaction detection.
- Personalized and safe medication recommendation under evolving health conditions.
In my talk, I will present our recent works, and talk about how we take deep learning approaches to address these challenges. In addition, I will also discuss several open challenges for applying deep learning in healthcare applications.
Bio: Cao (Danica) Xiao is a research staff member in the AI for Healthcare team at IBM Research.
Her research focuses on developing machine learning and deep learning models to solve real world healthcare challenges. Particularly, she is interested in deep computational phenotyping, adverse drug reaction signal detection from heterogeneous real world evidence, interpretable health analytics, and translational informatics research (e.g., drug similarity for drug safety and discovery).
The results of her research have been published in leading AI conferences including NIPS, ICLR, KDD, AAAI, IJCAI, SDM, ICDM, and top health informatics journals such as Nature Scientific Reports and JAMIA. She also serves as the principal investigator for several MIT-IBM joint projects. Prior to IBM Research, she acquired her Ph.D. degree from University of Washington, Seattle in 2016.