Artificial Intelligence for Risk Prediction and/or Stratification

• Activity Number: 34 - Current and Future Challenges in Diagnostic Device Evaluation
• Type: Topic Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 11:50 AM
• Sponsor: Section on Medical Devices and Diagnostics
• Author(s): Bipasa Biswas*
• Companies: Center for Devices and Radiological Health, FDA
• Keywords: Artificial intellegence ; probability score; diagnostic

Abstract

Artificial Intelligence (AI), is a technology that uses algorithm and software to combine large amount of data to learn automatically from patterns or features in the data and interpret underlined complex phenomena. AI is currently at the center of the medical horizon, expected to be used on an ongoing basis to change care pathways by expediting early detection and improve patient access to needed healthcare. Diagnostic devices utilizing Artificial Intelligence (AI), Deep Learning (DL) or machine learning (ML) have the potential to generate a risk score and/or probability of an outcome/event. This presentation will give an overview of the AI, DL and ML and discuss issues and challenges with probability scores.