Atlas‐Based Versus Individual‐Based Fiber Tracking of the Corpus Callosum in Patients with Multiple Sclerosis: Reliability and Clinical Correlations

Abstract

In multiple sclerosis (MS), the presence of lesions and normal-appearing white matter damage may affect the reliability of diffusion tensor (DT) magnetic resonance imaging (MRI)-based tractography. We compared the performance of an individual-based method for corpus callosum (CC) fiber tracking in MS with those of two atlas-based methods. Brain DT MRI scans were acquired from 35 patients with MS and 18 age-matched healthy volunteers (HV). DT-derived metrics from the CC-the mean diffusivity (MD) and fractional anisotropy (FA)-were calculated using an individual-based and two atlas-based methods with different types of subject registration (linear and nonlinear) to a CC atlas. Customized termination criteria were applied to stop the tracking algorithm when using the individual-based method. All the methods were able to distinguish between MS patients and HV. Using the individual-based method, stronger relationships were found between CC DT-derived metrics and the subjects' clinical condition. CC DT tractography using an individual-based method is more sensitive than the atlas-based ones to tract-specific alterations related to MS disability. An atlas-based method with nonlinear registration can be a valid alternative when an automated postprocessing is warranted, such as in the case of high volumes of data.