Date/Time : Thursday February 27, 11:45am - 1:00pm
Event Address: E62-450
Contact: (617) 324-7782 or
lmaloney@mit.eduTopic: Using advanced data sciences to empower clinical decision makers: inferring insulin regimen from clinical notes
Steven Spear, DBA MS MS is a senior lecturer at MIT
Monisha Pushpanathan is a Masters degree candidate in MIT’s Systems Design and Management Program
The focus of this talk is on augmenting the decision making of clinicians so they can more quickly arrive at tailored diagnoses and treatment plans for individual patients. Historically clinicians' capability reflected their individual experience only. Now data repositories contain enough collective experience, that, if tapped effectively, clinicians’ abilities could increase dramatically. We assess text mining techniques to infer past treatment decisions and results for Type 2 Diabetes insulin regimens by converting difficult to access information in unstructured clinical notes into a structured format using machine learning techniques, making it suitable for effective research and analysis.
This research is part of an ongoing effort to determine what insights can be acquired from written patient and caregiver recorded experience. Type 2 Diabetes was chosen as the target condition because it affects many patients, yet causality—the relationship between genotypical, phenotypical, and experiential characteristics and symptoms – is poorly mapped. If we can gain better insights into diabetes, other diseases—seizures, migraines, for example—invite similar investigation.
