Zhuang Song

Biography

I began to develop interest in brain science since my college time in China. After coming to the united states, I spent two years studying biological visual system in Vision Research Center at Vanderbilt University. In the first two years at Penn, I worked with Dr. Leif Finkel to study computational modeling of primary visual cortex and perceptual computing. Unfortunately I couldn't continue my work with Dr. Finkel because of the concern of his health. I then switched my thesis research to brain image analysis. The primary goal of my current research is to develop a robust and efficient statiscital segmentation methods to facilitate quantitative anaysis of neonatal brain images, which is essentail to investigate human brain development at critical early stage, in either normal or pathological states. Neonatal brain-MR image segmentation is a challenging problem due to low image-to-contrast ratio (CNR) but large biological variability in early brain development. The method under development can be also generalized to segment other medical images corrupted by severe noise, MRI inhomogeneity, and partial volume effect.

Projects

• Neonatal brain MRI segmentation
• Cerebellar tissue segmentation
• Graph-cut based segmentation
• Nonparametric statistical modeling and machine learning

Publications

Conference full papers (peer reviewed):

Conference Abstracts: