Medical Imaging AI: What the data tells us

Artificial Intelligence and Machine Learning provide the possibility of significantly enhancing how we practice radiology, now and in the future. While the technology holds great promise, there are many challenges to its successful deployment in clinical practice. Hundreds of algorithms have been cleared for use by the FDA, yet only a small minority of radiologists have deployed AI clinically. In this lecture, we will explore feedback on the use of AI from thousands of radiologists in a recent survey conducted by the American College of Radiology Data Science Institute, uncovering many challenges that exist in medical imaging AI today.

Learning objectives:
1. Recognize potential obstacles in deployment of AI into clinical setting
2. Identify specific challenges radiologists face using AI in medical imaging
3. Understand the FDA’s regulatory process for clearing algorithms for use

Rapid advances in artificial intelligence modelling have led to led to the development of new and exciting AI tools to assist radiologists. Developing a comprehensive chest X-ray diagnostic tool is a complex process, requiring consideration of clinical use cases, workflow integration and hidden stratification. Dr. Catherine Jones and Dr. Keith Dreyer discuss the process of developing and validating the most comprehensive chest X-ray diagnostic tool available.

Learning objectives:
1. To understand the commonly used terms relating to artificial intelligence in radiology.
2. To gain a deeper insight into the development of AI for the interpretation of chest radiographs.
3. To develop a greater understanding of the clinical use cases of chest radiograph AI.