Investigating subtle changes in tissue structure and metabolism is critical to diagnose diseases at early stages. Therefore, we have studied techniques to improve image resolution and contrast in various neuroimaging studies. For instance, we have recently obtained a very high resolution Diffusion Tensor Images (DTI) of the perforant path in the hippocampus, a very thin and curved axonal fiber tract. Apart from the small size, it resides in a region that is prone to severe distortions and signal loss due to variations in tissue magnetic susceptibility. This fiber tract is implicated in early stages of Alzheimer’s disease.
Pictured Above: Data from two subjects, S1 and S2. (a) and (b) show the coronal slice and the anatomical location of the perforant path. (c) shows the strength of diffusion signal perpendicular to the surface of entorhinal cortex. (d) illustrates diffusion signal strength perpendicular to the entorhinal cortex inside the section shown in (b).
Studies of brain development
Studies of brain development
We have studied age-associated changes in the cortex, subcortical structures and brain white matter tracts using high resolution T1 weighted images and Diffusion Tensor Images (DTI). The results revealed maps of brain development in preadolescent ages at very high spatial and temporal detail.
Pictured Above: Inflated brain surfaces with statistically significant reduction in cortical thickness with age shown with blue overlay (p<0.05, FDR corrected). Upper row is the lateral and medial surfaces of the left hemisphere and lower row is the right hemisphere. Each column shows changes in cortical thickness during the specified age range. The color bar indicates t-scores.
Pictured Above: Development of subcortical structures during preadolescent ages. Colors represent F-scores. Caudate, thalamus and hippocampus are shown inside the brain.
Pictured Above: Age-associated changes in brain white matter fibers. Significant increase in FA (left panel, red colors) and decrease in MD was shown in Corticospinal tracts. Similar changes were seen in most major fiber tracts in the brain during preadolescent ages.
Diffusion Kurtosis imaging and concussion
Diffusion MRI, in general, helps estimate the bulk behavior of water molecules’ random displacements inside tissues. A series of images were acquired from the brain with different diffusion weightings obtained by varying MRI gradient fields. Analysis of this series of brain images helps probe tissue microstructure noninvasively. The conventional approach called Diffusion Tensor Imaging (DTI) assumes a Gaussian distribution of water molecule displacements. However, in complex tissue structures, Gaussian distribution provides a crude estimate. This shortcoming is addressed by Diffusion Kurtosis Imaging (DKI), which estimates the distribution of water molecule displacements as a sum of a Gaussian component and a non-Gaussian component (Kurtosis). This requires an additional set of images acquired with a different set of diffusion weightings called “b-values”. So, conventional DTI acquires a series of diffusion weighted images with a single b-value, while DKI requires a series of diffusion weighted images with two different b-values. Then the post-processing software fits the Gaussian + Kurtosis model to estimate bulk behavior of water molecules more accurately. The output of this software provides both the conventional DTI parameters such as Fractional Anisotropy (FA) and Mean Diffusivity (MD) and also additional Kurtosis metrics depicting the complexity of tissues in better detail. Therefore, DKI is a clinically feasible extension of DTI. It examines the additional non-Gaussian diffusion effects that occur in the brain tissues, thereby overcoming the limitations of DTI. Below is the maps of conventional DTI maps (FA and MD) and kurtosis maps (Radial, Axial and Mean Kurtosis).
This technique was recently applied to a group of concussed athletes recruited for a study conducted by Michael McCrea, Ph.D. Widespread changes in mean diffusivity (MD) and focal changes in axial kurtosis in frontal brain regions were observed:
Pictured Above: Mean diffusivity decreased and axial kurtosis increased in mTBI subjects (p<0.05 corrected for multiple comparisons)