Meeting Banner
Abstract #1410

Automated Image Quality Assessment for Diffusion Tensor Data

James Meakin1, Serena Counsell2, Emer Hughes3, Jo V. Hajnal3, David J. Larkman3

1Physics Department, Imperial College London, London, UK; 2Imaging Sciences Department, Clinical Sciences Centre MRC, London, UK; 3Imaging Sciences Department, Imperial College London, London, UK

In single shot EPI based DTI motion between acquisitions can be corrected by image registration approaches but individually damaged images are generally uncorrectable. This work outlines an algorithm designed to automatically identify and reject damaged images in large diffusion datasets, an otherwise manually intensive step. The approach calculates an estimate of the diffusion images using a data driven forward model and then compares this prediction to the acquired data using 2D image correlation to identify outlier images. The method has been tested on 9 preterm infant data sets and 9 adult data sets.