events

executive

news

How You Measure Matters – Progress and Pitfalls of Machine Learning in Healthcare

Dr. Joseph Beals

Vice-President of Informatics and Innovation, PeraHealth, Inc.

October 30, 2019 15:30 - 17:00

Mackenzie Building Room 4359, Carleton University

Paid parking available on campus

 

abstract

The digitization of medical data and near universal adoption of electronic medical records in hospitals has ushered in the age of big data in healthcare. This has coincided with significant advances in the power and availability of machine learning and artificial intelligence tools. The rise of machine learning models in healthcare has been heralded by notable successes in areas requiring focused pattern recognition, such as in radiology, diagnosing diabetic retinopathy, and classifying cardiac arrhythmias. While the promise of machine learning in healthcare goes far beyond these applications, even the best researchers sometimes fall into the trap of using 'tools without rules' to the detriment of both their work and the wider field. This talk will highlight some of the common pitfalls underlying even the most sophisticated machine learning models. We will also discuss key considerations in the development of a general deterioration prediction model which has demonstrated real utility for clinicians providing care to hospitalized patients.

biography

Joseph Beals is Vice President of Informatics and Innovation at PeraHealth, Inc, where he leads the Analytics and Research team and oversees the integration of analytics functions across product development, account management and sales to drive innovation and growth. He has broad experience leading the commercialization of emerging technologies. Prior to joining PeraHealth in 2010 he worked at the Centre for Advanced Photonics and Electronics in the U.K., and at the Fraunhofer Center for Manufacturing Innovation in the U.S. Joseph has published articles in diverse fields, including medical predictive analytics, quantum electrodynamics, opto-electronics, neural-engineering, and fluid mechanics, and holds a patent in high-bandwidth optical data routing. He received his ScB from Brown University, an MBA from Carleton University, and a PhD from the University of Cambridge, where he was a National Science Foundation Research Fellow.