Meeting Banner
Abstract #2007

Reproducibility of Automated Measurements of Diffusion Tensor Imaging at 3T using Histogram Analysis

Ryan Hutten1, Shawn Sidharthan2, Christopher Glielmi2,3, Hongyan Du4, Fiona Malone2, Ann Ragin2,5, Robert Edelman2,5, Ying Wu2,5

1Radiology, Northshore University HealthSystem, Evanston , IL, United States; 2Radiology, Northshore University HealthSystem, Evanston, IL, United States; 3Siemens Healthcare, Chicago, IL, United States; 4Center for Clinical Research Informatics, Northshore University Health Systems, Evanston, IL, United States; 5Radiology, Feinberg School of Medicine, Chicago, IL, United States


Quantitative analysis of Diffusion Tensor Imaging (DTI) can be used to monitor and detect the onset of neurodegenerative diseases such as Alzheimer's and Parkinson's. This reliability study was conducted in order to show the reproducibility of an automated segmentation procedure incorporated with Fractional Anisotropy (FA) maps using several histogram metrics. When compared to the mean FA approach, the histogram median, mean, and peak location showed a higher reproducibility. These findings suggest that histogram metrics may be potentially more useful than the mean approach in the longitudinal studies of various chronic neurodegenerative diseases.