| Summary: | 碩士 === 國立中山大學 === 電機工程學系研究所 === 104 === Cerebral cortical thickness can provide important clinical information, helping the diagnosis of specific neurodegenerative diseases. In general, the automatic cortical thickness measurement methods can be classified into two types, surface-based and voxel-based. In this study, three methods, including one surface-based FreeSurfer and two voxel-based methods, DiReCT and Laplace-VBCT, are compared. A single subject was scanned for 24 times in a period of three and half months. The global and regional mean of cortical thickness are obtained for each experiment to exam the reproducibility. Results show that the coefficients of variance (CV) of three methods are very close to each other while FreeSurfer is found to have the smallest CV in most of DKT cortical regions than two other voxel-based methods.
In addition, magnetic resonance imaging experiments were performed on 15 young teenagers very-low-birth-weight born preterm (mean age 12.8 yr), all without brain injury, and age-matched 17 normal peers (mean age 13.8 yr) to assess the cerebral cortical thickness. The two-sample t-test with a p value of 0.05 is used to detect the statistical difference. Thicker cortical thickness was found in parietal, temporal, and occipital lobes of the preterm group by using all three measurement methods. In a sub-group comparison, thicker cortices are found in similar areas of young preterm compared with young control. However, when the old preterm is compared to the old control group, more cortical regions with thinner thickness are found than those with thicker cortex. In addition, the young preterm group has thinner cortex than old preterm, revealing the cortical thinning during childhood and early adolescence, which can be also observed but in much less area. All of our results suggest that the developmental cortical thinning of preterm born children is delayed to their term born peers, but the group deviation of cortical thickness is narrowed with increasing age.
|
|---|
