The Improvement of Breast Cancer Diagnosis Rate in Magnetic Resonance Imaging (MRI) using Fusion of Super Pixels and Fuzzy Connectedness

Precise segmentation of tumors in the breast is one of the most significant steps for MRIs and diagnosis tools using computers. Segmentation of the breast tumor is a demanding task due to some factors including partial volume effect, the similarity of the brightness of tumor texture with other surrounding non-tumor textures, variety in shape size and location of the tumor in different patients. Due to its vitality, the process of segmentation is carried out manually by specialists and its disadvantages are long computation time, and high cost. To overcome these issues, algorithms are required to segment images with high accuracy and no need for user intervention. This study presents a new method based on fuzzy connectedness algorithm and super pixels for tumor segmentation in magnetic resonance imaging (MRI). The proposed method is applied to a dataset built by the respected researchers on Matlab. The suggested method has been compared using two commonly used methods of clustering and morphological operators in tumor segmentation in magnetic resonance imaging (MRI). Mean average precision of 98.33 and the Dice similarity coefficient of 98.06 signifies the prominence of the suggested method in comparison with other methods compared using clustering algorithm 90.33 and morphological algorithm 91.83.