Diagnosis of Brain Tumor Using Combination of K-Means Clustering and Genetic Algorithm

Introduction: Medical image processing aimed at reducing human error rates attracted many researchers. The Segmentation of magnetic resonance image for tumor detection is one of the recognized challenges in the treatment of the disease. Considering the importance of this issue in the present study, the diagnosis of brain tumor is considered. Material and Methods: One of the most popular and most widely used methods in the field of segmentation of images of resonance imaging of the brain is the k-means clustering algorithm, which, despite the diagnosis of a tumor, fall in to local optimum problem, followed by a reduction in the accuracy of the diagnosis tumors are malignant. In this study, we aimed to solve this problem and subsequently increase the accuracy of diagnosis of malignant tumors, a GA-clustering combination of clustering based on k-means and genetic algorithms. Results: How to combine in the way that the genetic algorithm is applied to each repetition of the K-means algorithm and, by scanning more in the space of the answer, is trying to find higher quality cluster centers. The effectiveness of the proposed method has been investigated on a number of images of BRATS standard collections. It is also compared with the K-means algorithm. Conclusion: The results show that the proposed algorithm provides better results than the K-means algorithm.