Dr. Seong Mun delivered a presentation at the NVTC Virtual Impact AI Summit on June 15, 2020 titled “Artificial Intelligence in Radiology: Lessons Learned Over the Past 30 Years”.

Here is a brief synopsis of the presentation: The radiology community has been developing computer aided diagnosis (CAD) tools over the past 30 years before artificial intelligence hype. Recent publications reported that the commercial CAD products used for digital mammograms provided no statistically significant benefits over the performance of radiologists. Conversely, a CAD system for lung cancer screening showed a significant improvement in accuracy and efficiency. The research and investments are increasing on the use of AI tools for a wide range of radiology applications. In spite of the explosive amount of interest in the use of AI for medical imaging, and highly speculative stories in the press, clinical adoption and commercial success of these AI tools remains elusive. In this presentation we will highlight how imaging AI products are developed and tested for FDA. FDA approval however does not assure clinical and commercial success. We will summarize remaining issues of wider adoption of this potentially power technology.

A copy of his presentation is viewable here: