Detection of brain tumour in 2d MRI: implementation and critical review of clustering-based image segmentation methods

Abstract

Image segmentation can be defined as segregation or partitioning of images into multiple regions with the same predefined homogeneity criterion. Image segmentation is a crucial process in medical image analysis. This paper explores and investigates several unsupervised image segmentation approaches and their viability and performances in delineating tumour region in contrast enhanced T1-weighted brain MRI (Magnetic Resonance Imaging) scans. First and foremost, raw CE T1-weighted brain MR images are downloaded from a free online database. The images are then pre-processed and undergo an important process called skull stripping. Then, image segmentation techniques such as k-means clustering, Gaussian mixture model segmentation and fuzzy c-means are applied to the pre-processed MR images. The image segmentation results are evaluated using several performance measures, such as precision, recall, Tanimoto coefficient and Dice similarity index in reference to ground truth images. The highest average Dice coefficient is achieved by k-means (0.189) before post-processing and GMM (0.208) after post-processing. Unsupervised clustering-based brain tumour segmentation based on just image pixel intensity in single-spectral brain MRI without adaptive post-processing algorithm cannot achieve efficient and robust segmentation results.