AI in healthcare services - innovation with focus on business value

We discuss a machine learning product to decide which request for a medical procedure are clinically appropriate based on clinical information from the patient’s medical record provided by a human user. We describe the process to build and deploy this product within a low risk tolerance,non-technological corporate environment, and how focus on data-driven analysis, data quality and processes, and user experience can create incremental value in quick iterations and are a prerequisite for deriving value from machine learning at scale. We will also compare various deep architectures to identify whether free text from a patient’s clinical record provides supporting evidence for the clinical necessity of a given procedure.

Afik Gal is a healthcare executive, physician, Duke MBA and an entrepreneur with over 13 years of cross functional experience (technology, clinical and business) in healthcare analytics and digital health. Afik is currently a VP for technology innovation at eviCore healthcare after previously leading product innovation at QPIDHealth(acquired by eviCore). Prior to joining QPID, Afik led PWC’s healthcare innovation lab where he focused on business model innovation and matching technologies with payers, providers and pharma needs. As an entrepreneur Afik founded a boutique consulting firm which helped hospitals and tech vendors innovate their products and business models (client included Amdocs, SAP, Orbotech, Wix,NetApp). Afik also founded a clinical analytics startup and consulted multiple other digital healthcare companies.