Speaker "Sanji Fernando" Details Back



: Improving revenue cycle management with deep learning – a health care case study


Deep learning is already being applied in a variety of health care areas including training models to support clinical applications like disease prediction and diagnosis. But as promising clinical applications are, deep learning can have a bigger and more immediate impact on care quality and cost when they are applied to administrative and operational processes. For example, Optum recently trained and deployed a deep learning model to prioritize cases for manual review by clinicians. Most case reviews are challenging because evaluating treatment options against medical necessity criteria is subjective to the individual case manager’s judgment. In addition, there is a limit to how many cases a skilled clinician can review within a specific amount of time, which can result in denials, rework or inappropriate reimbursement of claims. However, deep learning works well for case reviews because there is a lot of labelled data available on the inputs and outputs of decisions from the review process enabling the model to better identify outliers. This allows all admissions to be screened and stratified so that those that need a second level review are prioritized saving health systems time, money and energy. In this session, Sanji Fernando will share his experience from the end-to-end process of building, deploying, and operating a deep learning model that improves hospital revenue cycle management. He will share key learnings on what worked well — data preparation, model selection, visual presentation of findings — and opportunities to expand and scale the use of this and other deep learning models in the future.
Who is this presentation for?
Professionals seeking to learn practical information from a real life example on developing and deploying a deep learning model in health care. There haven’t been many examples of deep learning models deployed in health care, making this session distinctive in content.
Prerequisite knowledge:

What you'll learn?
The audience will gain an understanding of how the presenter approached training a deep learning model, including use case understanding, performance metric selection, data preparation, tool selection and tradeoffs, model development and optimization, model selection, deployment and model operations.


Sanji Fernando is a senior vice president at Optum, where he leads the Artificial Intelligence (AI) and Analytics Platforms team. He is responsible for developing platforms that support the design and development of leading edge AI models and analytic tools for the enterprise. Previously, Sanji was a vice president at OptumLabs and led the OptumLabs Center for Applied Data Science (CADS). The CADS team applied breakthroughs in AI and machine learning to solve complex health care challenges for UnitedHealth Group (UHG) by developing and deploying software product concepts. CADS pioneered using deep learning to streamline administrative processes in revenue cycle management and developed graph analytics tools to support provider network design, among other innovations. Sanji joined OptumLabs in 2014 from Nokia, where he created Nokia’s first data science team. His team launched the first big data computing cluster at Nokia, using cluster derived insights on user activity and engagement to design new product concepts. Before that, Sanji spent 9 years at Nokia in a variety of corporate roles with Nokia’s Multimedia Division, Nokia Research Center and Nokia Ventures. Prior to Nokia, Sanji was a co-founder and VP of Engineering at Vettro, a venture-backed mobile software company. Sanji began his career in consulting with Viant and Accenture. Sanji is a graduate of Trinity College with a bachelor’s degree in computer science. He lives in the Boston area with his wife and their three boys. In his free time, Sanji enjoys coaching his sons in basketball and baseball. He also serves on the board of his local Little League.