ISMRM22 - Acquisition & Analysis in Context: Musculoskeletal Image Acquisition & Reconstruction

Joint Annual Meeting ISMRM-ESMRMB & ISMRT 31st Annual Meeting • 07-12 May 2022 • London, UK

Acquisition & Analysis in Context: Musculoskeletal Image Acquisition & Reconstruction

Sunrise Course

ORGANIZERS: Andrew Scott, Fang Liu, Mark Chiew, Khin Tha, Janine Lupo, Dan Ma, Brian Hargreaves

Wednesday, 11 May 2022

ICC Capital Suite 14-16

08:00 - 09:00

Moderators: Miika Nieminen

Skill Level: Basic to Intermediate

Session Number: S-W-06

Session Number: S-W-06

Overview
Junior Fellow Observer: Bragi Sveinsson

While many MR data acquisition and image analysis techniques are broadly applicable, there are often strong context-specific strategies that take into account the particular constraints, challenges, and opportunities that are unique to different clinical or physical contexts. This course will first review acquisition and analysis methods for clinical neuroimaging, followed by musculoskeletal imaging, and body imaging. A final session on acquisition and analysis in the context of different field strengths will be presented.

Target Audience
Clinicians, scientists, engineers, technologists, and other professionals interested in learning about acquisition and analysis techniques in different contexts.

Educational Objectives
As a result of attending this course, participants should be able to:
- Outline data acquisition and analysis methods used for clinical neuroimaging;
- Summarize the methods available for musculoskeletal imaging;
- Describe the acquisition and analysis strategies used in body imaging;
- Compare and contrast the advantages and disadvantages of ultra-high and low field MRI; and
- Recognize the differences in MR techniques used in different clinical and physical contexts.

	08:00		Rapid Musculoskeletal Imaging and Machine Learning: Acquisition, Reconstruction & Analysis View the Presentation Akshay Chaudhari Machine learning has enabled several advances for acceleration of MRI sequences along with the automated analysis of arbitrarily reconstructed images. This first portion of this lecture will focus on summarizing the major families of machine learning techniques applied towards undersampled image reconstruction tasks as well as the automation of classification, segmentation, and detection tasks. Since machine learning advances are strongly coupled to publicly available datasets, the second portion of this lecture will compare and contrast current public MRI data repositories relevant for musculoskeletal MRI.
	08:30		Imaging Osteoarthritis: Now & Future for Image Acquisition & Reconstruction? View the Presentation Hollis Potter Quantitative MRI of cartilage provides objective assessment of both collagen orientation and proteoglycan content. Ultrashort TE sequences provide robust assessment of collagen orientation in fibrocartilage such as the labrum and meniscus, as well as ligament and tendon collagen integrity. Quantitative MR techniques have shown to be reproducible across sites and have been correlated to altered gait mechanics. Voxel-based relaxometry allows for more facile assessment of longitudinal change, and the addition of convolutional neural networks permit assessment of large, curated databases to detect variables that lead to disease progression. Lastly, weightbearing assessment allows for determination of cartilage mechanics and scrutiny of surgical interventions aimed at delaying osteoarthritis progression.