Author:

Keywords:

Science & Technology, Life Sciences & Biomedicine, Clinical Neurology, Neuroimaging, Psychiatry, Neurosciences & Neurology, Diffusion MRI, Diffusion tensor imaging, Constrained spherical deconvolution, Tractography, White matter pathology, MICROSTRUCTURAL ABNORMALITIES, SPATIAL STATISTICS, TENSOR, BRAIN, MRI, PATHOPHYSIOLOGY, TRACTOGRAPHY, DEPRESSION, ANISOTROPY, INTEGRITY, Bipolar Disorder, Brain, Diffusion Magnetic Resonance Imaging, Diffusion Tensor Imaging, Humans, White Matter, 1103 Clinical Sciences, 1109 Neurosciences, 1702 Cognitive Sciences, 3202 Clinical sciences, 3209 Neurosciences, 5202 Biological psychology

Abstract:

White matter pathology likely contributes to the pathogenesis of bipolar disorder (BD). Most studies of white matter in BD have used diffusion tensor imaging (DTI), but the advent of more advanced multi-shell diffusion MRI imaging offers the possibility to investigate other aspects of white matter microstructure. Diffusion kurtosis imaging (DKI) extends the DTI model and provides additional measures related to diffusion restriction. Here, we investigated white matter in BD by applying whole-brain voxel-based analysis (VBA) and a network-based connectivity approach using constrained spherical deconvolution tractography to assess differences in DKI and DTI metrics between BD (n = 25) and controls (n = 24). The VBA showed lower mean kurtosis in the corona radiata and posterior association fibers in BD. Regional differences in connectivity were indicated by lower mean kurtosis and kurtosis anisotropy in streamlines traversing the temporal and occipital lobes, and lower mean axial kurtosis in the right cerebellar, thalamo-subcortical pathways in BD. Significant differences were not seen in DTI metrics following FDR-correction. The DKI findings indicate altered connectivity across cortical, subcortical and cerebellar areas in BD. DKI is sensitive to different microstructural properties and is a useful complementary technique to DTI to more fully investigate white matter in BD.