A Heat Kernel Based Cortical Thickness Estimation Algorithm

Cortical thickness estimation in magnetic resonance imaging (MRI) is an important technique for research on brain development and neurodegenerative diseases. This paper presents a heat kernel based cortical thickness estimation algorithm, which is driven by the graph spectrum and the heat kernel theory, to capture the grey matter geometry information in the in vivo brain MR images. First, we use the harmonic energy function to establish the tetrahedral mesh matching with the MR images and generate the Laplace-Beltrami operator matrix which includes the inherent geometric characteristics of the tetrahedral mesh. Second, the isothermal surfaces are computed by the finite element method with the volumetric Laplace-Beltrami operator and the direction of the steamline is obtained by tracing the maximum heat transfer probability based on the heat kernel diffusion. Thereby we can calculate the cerebral cortex thickness information between the point on the outer surface and the corresponding point on the inner surface. The method relies on intrinsic brain geometry structure and the computation is robust and accurate. To validate our algorithm, we apply it to study the thickness differences associated with Alzheimer’s disease (AD) and mild cognitive impairment (MCI) on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset. Our preliminary experimental results in 151 subjects (51 AD, 45 MCI, 55 controls) show that the new algorithm successfully detects statistically significant difference among patients of AD, MCI and healthy control subjects. The results also indicate that the new method may have better performance than the Freesurfer software. keywords: Cortical thickness, Heat Kernel, Tetrahedral Mesh, Streamline, False Discovery Rate