An Adaptive Level Set Evolution for the Analysis of Ventricle Variations in Alzheimer MR Images

Alzheimer-s disease (AD) is a common form of dementia and it is fatal neurogenerative disorder resulting in atrophy of brain regions. MR imaging is a very important tool in diagnosing AD. The enlargement of ventricles due to neuronal loss is a significant characteristic of AD. In this work, ventricular expansion in Axial, Saggital and Coronal views of MR images is analyzed by segmenting the ventricles. Due to intensity inhomogeneties segmentation techniques are to be more sensitive to capture variations in the structural boundaries. Intensity inhomogeneties may cause considerable difficulties in image segmentation. In order to overcome this difficulty we propose a region based active contour model that draws upon intensity information in local regions at a controllable scale. A data fitting energy is defined in terms of a contour and this energy is incorporated into level set formulation. Due to kernel function intensity information in local regions is extracted to guide the motion of the contour. The size and shape of the ventricles in AD subjects are apparently enlarged in the axial view than in the sagittal and coronal views. Segmentation of dilated ventricle is a key step in the diagnosis of AD.