Using Machine Intelligence to Reduce Clinical Variation
February 21, 2017 — 01:00PM EST - 02:00PM EST
Orange County Convention Center
Using a combination of machine learning and an application of mathematics called topological data analysis (TDA), hospitals, health systems, and integrated delivery networks can accelerate the analysis of large and complex datasets with “unsupervised” discovery, capturing insights faster and more comprehensively than traditional or homegrown analytical tools. This unbiased approach of machine intelligence for pattern identification unlocks insight into practice variation for clinical pathways and population management.
The speakers will illustrate the application of machine intelligence for optimizing care for total joint replacement and laparoscopic surgery patients at Mercy Health System, using EHR data from 10 hospitals. The speakers also will demonstrate the use of machine intelligence for analyzing health claims data for chronic disease population segmentation and identify opportunities to advance precision medicine.
Contrast traditional, hypothesis-driven inquiry with “unsupervised” learning using machine learning and topological data analysis (TDA)
Outline strengths and weaknesses of large datasets from EHRS, claims and genetic analysis related to variation and anomaly detection
Demonstrate the use of topological data analysis for surgical pathway optimization
Demonstrate new methods for understanding illness patterns in chronic disease populations using machine learning for claims data study and biomarker analysis