Open-Source Pipelines for MSK Applications
Sunrise Course
ORGANIZERS: Feliks Kogan, Fang Liu
Thursday, 09 May 2024
Summit 1
07:00 -
08:00
Moderators: Albert Woo Ju Jang & Rianne van der Heijden
Skill Level: Intermediate to Advanced
Session Number: S-Th-01
CME Credit
Session Number: S-Th-01
Overview
Quantitative MRI (qMRI) in the musculoskeletal system has shown tremendous potential but has been hampered by a fragmented landscape for analysis of qMRI data as well as a lack of easily accessible analysis code for non-physicists. Open-source analysis methods such as freesurfer, FSL, and other analysis packages have greatly benefited and advanced their field in both the technical MRI community and beyond.
This session will present an overview of some open-source software that is being developed for the MSK MRI community, particularly for (1) whole-joint and (2) muscle imaging. Further, there will be a presentation and discussion on how to maximize the benefit and utility of such software for the community as a whole.
Target Audience
Clinicians, scientists, and engineers interested in musculoskeletal MRI.
Educational Objectives
As a result of attending this course, participants should be able to:
- Summarize the current status of open-source analysis pipelines in MSK imaging including the roles of automated segmentation, quantitative parameter fitting, and outcomes analysis;
- Discuss how the MRI community can work to build on these efforts together to benefit the field. This includes modular platforms that can assess and compare multiple analysis approaches; and
- Identify how open-source platforms have enhanced other areas within ISMRM and what is needed as well as the challenges that face MSK MRI open-source technologies.
| 07:00 | | Open-Source Muscle Analysis Francesco Santini Keywords: Musculoskeletal: Muscular, Transferable skills: Reproducible research In the MRI of muscular and neuromuscular diseases, reproducibility of the pipeline is of paramount importance. Usually, these diseases are so rare that single-center studies are underpowered to capture the variability of the population, but the step to effective multicenter trials passes through the availability of standardized and open methods, which can be implemented on a variety of machines and data. In this talk, I will present some of the current solutions, and the current challenges that are being faced to build a completely reproducible muscle MR workflow. |
| 07:20 | | Adopting Open-Source Deep Learning for MSK Workflows Arjun Desai Keywords: Transferable skills: Reproducible research, Musculoskeletal: Knee, Transferable skills: Software engineering Deep learning (DL) has shown promise for a variety of applications in MRI from upstream acquisition and reconstruction to downstream image analysis. However, while new DL models for MRI are being developed and open-sourced at unprecedented rates, their translation into tools that can be used by research and clinical practitioners has been limited. In this talk, we will discuss the challenges, solutions and opportunities for building user-centric, DL-assisted tools for MSK MRI workflows. |
| 07:40 |  | Open-Source Challenges & Opportunities Mark Chiew Keywords: Transferable skills: Software engineering, Transferable skills: Reproducible research In a research field that is heavily dependent on software for image reconstruction, processing and analysis, open source software is a key component of the MR research environment. In this talk, we will discuss the opportunities as well as some of the associated challenges of open source software development in MSK and MRI research in general. |