ISMRM21 - The Future Is Now: Artificial Intelligence for Body Imaging

The Future Is Now: Artificial Intelligence for Body Imaging

Sunrise Session

ORGANIZERS: Daniel Margolis, Johannes Heverhagen, Mustafa Shadi Bashir

Tuesday, 18 May 2021

Concurrent 7

13:00 - 14:00

Moderators: Piotr Kozlowski & KyungHyun Sung

Skill Level: Basic to Advanced

Session Number: S-02d

Parent Session: Hot Topics in Body MRI

Session Number: S-02d

Overview
Debate whether artificial intelligence (AI) is "ready for prime time," where it currently adds value, and where it falls short.

Target Audience
Practicing clinicians and artificial intelligence developers.

Educational Objectives
As a result of attending this course, participants should be able to:
- Identify at least one area where artificial intelligence adds value;
- Discriminate between the hype and reality of AI; and
- Critically evaluate new innovations as they become available.

		Artificial Intelligence: Fits & Spurts Watch the Video Masoom Haider In this presentation we will review whether artificial intelligence (AI) is "ready for prime time," where it currently adds value, and where it falls short. Specifically the state of affairs in 2021 of Body MRI AI applications will be reviewed. The needed steps for translation of AI tools into clinical practice will be discussed including need for data repositories, control of data bias, performance of prostate MRI CAD, need for robustness and the status of FDA approval for AI.
		Artificial Intelligence: Smooth Sailing Watch the Video Silvia Chang Rapid advancement of artificial intelligence (AI) in Radiology is due to increasing availability of datasets, stronger computing power and improved learning algorithms. This area of active investigation already has some algorithms FDA approved for clinical practice. AI can augment radiologists’ workload, thereby improving their efficiency and accuracy. This includes triaging, segmentation, detection and classification of tumor aggressiveness, report generation and prognostication. AI will relieve the radiologists from repetitive mundane tasks and enable her/him to focus on tasks that AI cannot do such as perform image-guided procedures. All this will result in decreased healthcare costs and most importantly improve patient outcomes.