Adaptive Technique Depending on Region Growing and Soft Clustering to Detect Tumors in Different Modalities of MRI Brain Images

Abstract

Brain tumor is a very dangerous disease and life threatening, so early detection of the tumor is a vital task. Many techniques and algorithms are presented to enable doctors for fast and accurate diagnosis of tumors in MRI brain images. In this study, analytical study of Region Growing segmentation method with different threshold values ranged from 10 to 35, with steps of 5, is proposed. In addition, an adaptive technique is proposed, which is Region Growing based on the fuzzy clustering scheme to investigate the performance of this algorithm by implementing it on FCM clustered images. The adopted MRI images are of different modalities and different orientations to test the ability of the adaptive technique to segment different modalities of MRI images. The results showed that, utilizing different values of the threshold in proceeding of Region Growing algorithm produced different segmented images’ properties. When the fine details of the processed images and their objects are the goal, low values of threshold must be adopted, while when isolating of the hole tumor regions is the goal, high values of threshold must be adopted. In addition, the results of the adaptive technique showed that Region Growing segmentation improved its performance and it could separate the consists of the tumor regions. The elapsed time of implementation is clearly reduced.