Abstract #3032

Using A Hyperspherical Harmonic Basis for Sparse Multi-Voxel Modeling of Diffusion MRI

Evan Schwab^1,2, Hasan Ertan Cetingul², Rene Vidal³, and Mariappan Nadar²

¹Electrical and Computer Engineering, Johns Hopkins University, Baltimore, MD, United States, ²Medical Imaging Technologies, Siemens Healthcare, Princeton, NJ, United States, ³Biomedical Engineering, Johns Hopkins University, Baltimore, MD, United States

For diffusion magnetic resonance imaging (dMRI), 3D signals are acquired at each voxel to estimate neuronal fiber orientation in the brain. Traditionally, dMRI signals are reconstructed using the same basis for each voxel with added spatial regularization and sparsity constraints. By repeating the same basis for each voxel, there exist millions of redundant parameters to represent an entire brain volume. In this work, we reconstruct dMRI signals jointly across multiple voxels to reduce the number of parameters needed to represent a brain volume by 65%. Sparse dMRI representation is important for storage, information extraction, and reduction of clinical scan times.

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

Join Here