ISMRM21 - AI-Augmented Musculoskeletal MRI

AI-Augmented Musculoskeletal MRI

Sunrise Session

ORGANIZERS: Jan Fritz, Kimberly Amrami, Hiroshi Yoshioka, Edwin Oei, Xiaojuan Li

Wednesday, 19 May 2021

Concurrent 7

13:00 - 14:00

Moderators: Akshay Chaudhari & Young Han Lee

Skill Level: Basic to Intermediate

Session Number: S-03b

Parent Session: Hot Topics in MSK MRI

Session Number: S-03b

Overview
This course will review and discuss state-of-the-art acquisition, reconstruction and post-processing techniques of MSK MRI, as well as the applications of AI techniques in MSK MRI and how it improves the clinical practice in MSK radiology.

Target Audience
Radiologists, clinicians, technologists, and scientists interested in learning about acquisition, reconstruction, and post-processing techniques and AI applications in MSK MRI.

Educational Objectives
As a result of attending this course, participants should be able to:
- Describe the state-of-the-art acquisition, reconstruction, and post-processing techniques of MSK MRI;
- Describe the applications of AI technologies in MSK MRI; and
- Discuss the potential clinical impact that AI technologies may bring to MSK radiology.

		AI-Based Acquisition, Reconstruction & Post-Processing Watch the Video Joshua Trzasko Musculoskeletal (MSK) imaging is a complex area of MRI with unique goals, requirements, and challenges. In recent years, Artificial Intelligence (AI) technologies – including machine (ML) and deep learning (DL) – have facilitated many new capabilities in this field that were previously impossible with conventional imaging technologies. This talk will overview the technical mechanics of AI/ML and discuss how techniques from this field can be applied for MSK MRI, including fast imaging, denoising, superresolution, motion correction, parameter mapping, segmentation, and radiomics. Example applications from recent literature as well as emerging directions of AI/ML technologies for MSK MRI will be discussed.
		The AI-Augmented MSK Radiologist Watch the Video Hollis Potter AI is rapidly being adopted into orthopaedic imaging, improving accuracy and speed of acquisition, improving image presentation and reproducibility, and aiding in image interpretation including fracture detection, osteoarthritis metrics and degenerative spine changes. Other applications include cartilage and meniscal segmentation, more efficient interpretation of large quantities of quantitative MR data, and the ability to increase spatial resolution and disease detection. Examples of pitfalls in study design are presented, as well as current applications of AI into both conventional and advanced imaging.