Open and reproducible pipeline for the acquisition and analysis of muscle MRI data in Facioscapulohumeral Muscular Dystrophy

Dr. Claudia Weidensteiner, University Hospital Basel

Abstract

Facioscapulohumeral muscular dystrophy (FSHD) is a rare neuromuscular disease characterized by muscle degeneration and fat infiltration leading to progressive muscle weakness. MRI scans, especially for the measurement of the extent of the fat infiltration in the muscles, are crucial for the diagnosis and monitoring of FSHD. Only recently, after identifying the primary genetic origin of the disease, targeted, effective drug candidates are entering clinical development. For upcoming clinical multicenter studies, standardized, easy-to-use data acquisition and post-processing tools have to be available that are capable of acquiring and processing MRI data on/from a range of scanners. The necessity for clinical trial readiness was expressed in several workshops and papers in the muscular imaging community. In this project, we will create a fully reproducible, vendor-independent solution for the acquisition of MR biomarkers of FSHD, from image formation to postprocessing, by refining existing open-source tools and combining them with new solutions. For data acquisition, we will develop a scanner-independent acquisition sequence based on multi-echo spin-echo (MESE) sequences for the measurement of the T2 of water, which is an indicator of edema and acute disease activity. For data analysis, we will adapt the existing open-source tool for water T2, and will implement a fat fraction calculation from the most commonly used 2-point-Dixon fat/water measurements. The pipeline will offer standardized input/output and tools for anonymization and segmentation. Testing of the post-processing tools will be performed on existing data of clinical FSHD studies (patients and healthy volunteers). All software development will be done in Python, a high-level programming language that is free and open-source and has vast library support. The goal of our open source software solution is to advance independent, reproducible academic research especially in (but not limited to) the field of FSHD.