Region based active contours model with contour let transform for segmentation of brain tumor from MR images

Abstract

Brain tumor segmentation is a routine process in a clinical setting and provides useful information for diagnosis and treatment planning. Manual segmentation, performed by physicians or radiologists, is a time-consuming task due to the large
quantity of medical data generated presently. Hence, automatic segmentation methods are needed and several approaches
have been introduced in recent years including region based active contour models. Region-based active contour models
are effective in segmenting images with poorly defined boundaries but often fail when applied to images containing intensity inhomogeneity. A new multiscale and multidirectional technique called contourlets transform is used to captures edges
and smooth contours at any orientation for noise removal and enhancement for the image superiority. The traditional models utilize pixel intensity and are very sensitive to parameter tuning. In this paper, a new region-based active contour model,
namely local region-based Chan–Vese (LRCV) model with contourlet transform is proposed for the segmentation brain
tumor from MR images. By considering the image local characteristics, the proposed model can effectively and efficiently
segment images with intensity inhomogeneity. An experimental result on real MR images shows an improvement in terms
segmentation accuracy and robustness towards noisy images. Compared with the existing techniques, the proposed model
is computationally more efficient and convergence of contour towards the tumor boundary is very accurate.