Automated Brain Image Classification Using Nature-Inspired Optimization-Based Machine Learning Algorithm

The field of medicine greatly depends on medical imaging. In modern brain imaging, image classification is used to separate abnormal tissues from healthy tissue. Using various categorization algorithms, the region of the brain tumor as well as the brain tumor size is recognized in the MRI images, allowing the brain tumor to be diagnosed. Brain Magnetic resonance imaging (MRI) can quickly and effectively identify tumors, aiding neurologists in their diagnosis. Cancer, the most prevalent or leading cause of death globally, may be made more likely by tumors. Currently, efficient automation of tumor detection is crucial to finding brain tumors. In this research we suggested a strategy that combines the strength of machine learning algorithms with nature-inspired optimization approaches to classify brain image automatically. The study proposed a new classifier called Self-Adaptive Deer Hunting Optimized k-Nearest Neighbors (S-ADHO-k-NN). To start, the input image is first preprocessed to eliminate any noise using the median filtering (MF) method. The image was classified using a suggested S-ADHO-k-NN. The experimental results showed that the S-ADHO-k-NN model significantly outperformed the comparative methods with regard to of effectiveness.