ASGE Masterclass: Artificial Intelligence (On-Demand) | June 2021-AI in predictive modeling: Complex EHR Data

Video Transcription

Video Summary

In this video, TAPOCHE Roy, a senior manager at Kaiser Permanente, discusses the role of computer vision and endoscopy in analyzing big data sets and predictive modeling in healthcare. TAPOCHE explains that electronic health records (EHR) play a crucial role in capturing and analyzing data in healthcare. The goal of artificial intelligence (AI) in EHR is to provide insights at both the individual and population level. The data in healthcare is heterogeneous, including structured text, unstructured data, images, video, and audio. TAPOCHE talks about the complexities of healthcare data and mentions a framework by Dr. Thomas Kannapalli for understanding complexity in healthcare systems. He then discusses the life cycle of predictive modeling, including data cleaning, feature engineering, model development and deployment, verification, monitoring, and taking actions. TAPOCHE explains the paradigms of machine learning - unsupervised learning, supervised learning, and reinforcement learning. He gives examples of classification, regression, and clustering models used in healthcare. TAPOCHE also discusses opportunities for improvement in machine learning, such as fairness, bias, transparency, explainability, and trustability. He concludes by presenting an EHR solution for population and individual views.

Asset Subtitle

Taposh Roy

Keywords

computer vision

predictive modeling

electronic health records

machine learning paradigms

EHR solution