Wavelet-based image registration and segmentation framework for the quantitative evaluation of hydrocephalus

Abstract

Hydrocephalus, a condition of increased fluid in the brain, is traditionally diagnosed by a visual assessment of CT scans. This thesis developed a quantitative measure of the change in ventricular volume over time. The framework includes: adaptive registration based on mutual information and wavelet multiresolution analysis, adaptive segmentation with a novel feature extraction method based on Dual-Tree Complex Wavelet Transform (DT-CWT) coefficients, and a volume calculation. The framework, when tested on physical phantoms had volume calculation accuracy of 1.0%. When tested on 8 clinical cases, the results reflected and predicted the diagnosis of the doctors, with less than 5% calculated volume change for cases where the diagnosis indicated the patient was stable, and more than 20% calculated volume change for cases for which hydrocephalus had been diagnosed. The outcome illustrated that the framework has good potential for development as a tool to aid in the diagnosis of hydrocephalus.