Data Sharing

July 6 – 7, 2019 | 5:00pm – 8:00pm | Washington D.C.
Add to Calendar

Access to data is crucial for the development of AI. Yet before safely sharing data, physicians face challenges, including defining who owns it, determining its value and evaluating whether it should be shared. In this video series, we’ll explore the top considerations and look at how physicians can use data pooling and federated learning to create more generalizable AI models.

Data Sharing from a Physician's Perspective

By: Ross Filice, MD



Most AI models in medical imaging come from a single institution. This type of model typically doesn’t generalize well to other locations due to limited sample size, rare diseases or the use of different equipment. This video outlines how physicians can use data pooling and federated learning to share data and create more generalizable AI models.

Learning Objectives:
  • Understand why physicians should care about data sharing.
  • Discover two methods to improve generalizability of AI models.
  • Grasp the pros and cons of data pooling.
  • Learn the pros and cons of federated learning.



An IT Member's Perspective on Data Sharing

By: Brian Baker, BS



This video provides an overview of data sharing, from the perspective of an IT member at a teleradiology practice, for the purposes of building machine learning models. This particular practice uses a five-tiered approach to safe data sharing, including Business Associate Agreements, image de-identification, and data encryption.

Learning Objectives
  • Understand a five-tiered approach to data sharing.
  • Learn best practices for IT members to safely share data.






Patient Perspective on Privacy & Data Sharing

By: Andrea Borondy-Kitts, MS, MPH



When sharing patient data, physicians should focus on building trust and transparency with their patients. When these are the foundation, most patients express willingness to share some data. This video reviews top concerns and considerations patients have when sharing their data for AI purposes and how data sharing agreements, fair compensation, and patient medical record apps could increase participation in data sharing.

Learning Objectives:
  • Understand key patient considerations and concerns when sharing data for AI.
  • Identify new trends likely to impact data sharing.
  • Learn how to increase patient participation in data sharing.
  • Discover different sources of digital health-related data.
  • Know what to include in data sharing agreements with patients.

A Lawyer’s Perspective on Data Sharing

By: Tom Hoffman, JD, CAE



There are many legal implications to consider when obtaining and sharing patient data. This video reviews the top challenges and opportunities physicians face when sharing data, including defining who owns it, determining its value, strategizing about how it can be monetized, and considering whether or not it should be shared.

Learning Objectives:
  • Comprehend the top challenges and opportunities with data sharing and privacy.
  • Learn the importance of defining who owns patient data.
  • Understand why physicians might not share data, even if they have the legal right to do so.