Automated Brain Tumor Segmentation for MR Brain Images Using Artificial Bee Colony Combined With Interval Type-II Fuzzy Technique

Abstract

Accurate prediction of brain tumors is vital while getting to the forum of medical image analysis, where precision in decision-making is of paramount importance, and the problems are to be addressed forthwith. For over a decade, innumerable medical imaging techniques using artificial intelligence and machine learning have been promulgated. The present research is intended to develop an algorithm that forges the working principles of the Artificial Bee Colony (ABC) and Interval Type-II fuzzy logic system (IT2FLS) algorithm to delineate the tumor region, which has been encompassed by complex brain tissues. The crux of any therapeutic sequences to be accomplished lies in the decisiveness of the oncologists, where the algorithm presented in this study significantly leverages decision-making through technological intervention. The algorithm proposed has versatility in handling a wide range of image sequences available in the BRATS challenge datasets (2015, 2017, and 2018) that have various levels of barriers, setbacks and hardships in identifying the aberrant regions, and it provides better segmentation outcomes that have been qualitatively validated and justified with metrics, such as DOI, specificity and sensitivity. Augmentation of the visual perception for oncologists is the insignia of this study, which in turn provides better insight and understanding regarding the ailment of the patient.