Automated quantitative analysis of human MR head images and study of age-related human brain volume changes

In this study, algorithms for automated analysis of human NM head images were developed and their application to quantification of age-related human brain volume changes was addressed.; First, we developed a histogram-based brain segmentation (HBRS) algorithm based on histograms and simple morphological operations. The algorithm's three steps are foreground/background thresholding; disconnection of brain from skull; and removal of residue fragments (sinus, cerebrospinal fluid, dura and marrow). Systematic evaluations of the algorithm indicated that the accuracy of the algorithm is higher than 98% and the reproducibility greater than 99%. These results showed that the HBRS algorithm is a simple, fast, and accurate method to determine brain volume with high accuracy and reproducibility.; Second, we described a novel knowledge-based automated method to identify brain sulci, and then defined the hemispheres and frontal lobes. The algorithms identify cortical landmarks directly instead of resorting to warping procedures. The medial surfaces of the sulci were extracted and identified; and brain images were divided into the left and right hemispheres and frontal lobes based on these medial surfaces. The accuracy was evaluated by both of the comparison to results of manual tracing and visual inspection. The results suggested that the method might be used to directly identify human brain structures with high accuracy.; Third, we measured volumes of the whole brain, left and right hemispheres, and frontal lobes from young and elderly individuals using the automated methods described above. We found that brain volume decreased with age; and the decrease was more pronounced in the frontal lobes. The coefficients of variances (CV) of the whole and sub-brain structure volumes were smaller in the old than in the young group. We found that the left hemisphere volume decreased more than that of the right hemisphere, but there is no such age-related left-right asymmetry in the frontal lobes. The lower CV values of volumes of the whole brain, hemispheres, and frontal lobes in the old people suggested that the ontogenetic constraints in old people might be more pronounced than that in the young people. The left-right asymmetry analysis suggested that the cerebral function reorganization happened not only in individual functional areas, but also at the system level.