Gastroenterology and Artificial Intelligence: 4th Annual Global Artificial Intelligence Summit | August 2022-Learning from Other Disciplines: How Far Behind is GI?

Video Transcription

Video Summary

In this video, Dr. Charles Kahn, a professor and radiologist at the University of Pennsylvania, discusses the challenges and considerations regarding artificial intelligence (AI) in medical imaging. He highlights the importance of learning from other specialties and addresses the question of how far behind gastroenterology (GI) is in adopting AI. Dr. Kahn emphasizes the need for rigorous testing and validation of AI systems, particularly in terms of establishing ground truth and evaluating the impact of human bias. He also discusses the importance of information standards, such as DICOM, in enabling interoperability and innovation in radiology. Dr. Kahn points out the challenges of external validation and the need to test AI systems on specific populations. He also discusses the potential bias and ethical concerns that can arise from AI systems, including associations and shortcut learning. Lastly, he highlights the importance of comprehensive evaluation and guidelines when considering commercial AI solutions. At the conclusion of his talk, Dr. Kahn expresses his appreciation for the shared challenges and emphasizes the need for collaboration in addressing them.

Asset Subtitle

Charles Kahn, Jr., MD, MS

Keywords

artificial intelligence

medical imaging

challenges

validation

interoperability