Meeting Banner
Abstract #1402

Multi-parametric MRI-based classification for generating muscle percentage index in muscular dystrophy

Aydin Eresen1, Noor E. Hafsa2, Lejla Alic2, Sharla M. Birch1, Jay F. Griffin1, Joe N. Kornegay1, and Jim X. Ji1,2

1Texas A&M University, College Station, TX, United States, 2Texas A&M University at Qatar, Doha, Qatar

Imaging biomarker for muscular dystrophies, such as muscle percentage index (MPI), successfully differentiates between healthy and dystrophic muscles. However, the current methods to generate this biomarker are not well defined and therefore lack robustness and reproducibility. This study imaged ten Golden Retriever Muscular Dystrophy (GRMD) pectineus-muscle samples at a 4.7T MRI scanner. To facilitate estimation of MPI and to validate the results, we use trichrome-stained histology images. These images were registered accurately to multi-parametric quantitative MRI (qMRI). We use local gradient and texture information to classify qMRI into muscle and non-muscle with respective accuracies of 0.86 and 0.71.

This abstract and the presentation materials are available to members only; a login is required.

Join Here