Short Time Optimized Super Pixels Segmentation for Effective Detection of Astrocytoma Brain Tumors

Clinical scanning serves an essential part in the diagnostic procedure of several severe disorders and the following medication procedures of a patient. The brain is a major and very complicated component in the internal organs that functions through billions of neurons, that regulates all the other organ’s operations. Because the brain is a fragile, intricate, and crucial part of the human body, it is one of the most common causes of death among cancer patients. However, a good and prompt treatment may save lives to a certain degree. Hence, in this publication, an effective brain tumor identification framework is suggested utilizing Deformable model of Fuzzy C-Mean clustering (DMFCM), Adaptive Cluster with Super Pixel Segmentation (ACSP) and Gray Wolf Optimization with Adaptive Clustering with Super pixel Segmentation (GWO_ACSP) and are mainly tested on CANCER IMAGE ACHRCHIEVE (CIA) which is a database containing High Grade and Low-Grade astrocytoma tumor images and also with BRATS 2015. The evaluation matrices Accuracy, Dice coefficient, Jaccard score, MCC, True positive Rate (Sensitivity), Specificity, Positive Predictive Rate (Precision), F Score and Recall were computed in which the proposed Gray Wolf Optimization-based ACSP (GWO_ACSP) gives a better answer for brain tumor segmentation with an accuracy of 0.99% than other models like RG, PFCM, SLPSO, MRG. The wolf algorithm is more readily coupled with realistic design issues owing to its benefits of simplicity in concept, rapid searching rate, good search accuracy, and ease of implementation.