Posted by Friends of FSH Research on May 18, 2024
Report by Seth D Friedman
See also Kennedy Krieger Institute Data-Lake Expansion
The progress report highlights advancements in constructing an MRI data-lake for studying Facioscapulohumeral Muscular Dystrophy (FSHD) using AI segmentation techniques. Key accomplishments include demonstrating the effectiveness of AI in creating individual muscle maps from FSHD MRI datasets and integrating data from multiple site longitudinal studies into the repository. This integration has facilitated the optimization of AI techniques for analyzing non-standard images and the development of methods for assessing individual muscle scan coverage and segmentation. Improved reliability of fat fraction measurements and the introduction of linear and pixel-based measures for visualizing fat distribution have enhanced the accuracy and insight of data analysis. Efforts have also been directed towards exploring the relationships between MRI measures and functional data, laying the groundwork for understanding how structural changes correlate with functional impairment.
Moving forward, the focus will be on developing multivariate models for complex task performance and investigating the underlying mechanisms of fat deposition in individual muscles. Collaboration aims to develop a Bayesian multivariate disease progression model to derive personalized disease progression rates and refinement of methods toward further refinement of muscle segmental anatomy and how these boundaries constrain fat increases and functional deficits. These efforts are geared towards refining treatment strategies and improving our general understanding of individual muscle features in FSHD.
Connect with us on social media