A new fast Chinese Visible Human brain skull stripping method
Yunjie Chen, Zhang Jianwei, Wang Shunfeng · 2009
Image data of the entire cadaver from the Chinese visible human is being used to produce three-dimensional images and software for humanhuman anatomy research. The anatomy of human brain is more complicated. The Chinese Visible Human data was serially sectioned at 0.2 mm intervals without anything perfused. Each sectioned surface was inputted into a personal computer to produce anatomical images. With the effect of noise, bias field, and fake grey matters, it is a challenging task to build a digital three dimensional representation of a human brain. In order to obtain more accurate representation, the brain regions must be separated from highly variable background regions to obtain a suitable stack of segmentation images. We use adapted Gauss Mixture model as a promising starting point for a sophisticated segmentation framework of color images within 3-dimensions. The model can classify images meanwhile estimate the bias field. For the effect of the fake grey matters, a proper image preprocessing strategy turned out to be necessary for accurate and robust segmentation results. We present a complete high resolution and accurate segmentation of the CVH brain. Based on these images, 3D representation is presented.