Abstract #2177

Automatic Segmentation of White Matter Hyperintensities Based on Reaction Diffusion with Adaptive Threshold

Shuangxi Ji¹, Yining Huang², Jing Fang^1,3, Jue Zhang^1,3

¹College of Engineering, Peking University, BEIJING, China, People's Republic of; ²Dept.of Neurology, Peking University, BEIJING, China, People's Republic of; ³Academy for Advanced Interdisciplinary Studies, Peking University, BEIJING, China, People's Republic of

Quantitative analysis of white matter hyperintensities (WMH) is critical for identifying risk factors and understanding pathogenesis. In this retrospective study, a novel MR image processing strategy based on the modified diffusion reaction model with adaptive threshold was proposed to segment automatically WMH MR Images. T2FLAIAR images of ten subjects aging from 52 to 75 years old were used for evaluating the validation of the proposed segmentation strategy. The segmented results had a good agreement comparing with the size and location of the T2FLAIR images, indicating the proposed approach could offer more convenient tool for WMH quantitative analysis.