A Novel Supervised Learning Model for MRI Image Using Fuzzy Clustering and Naïve Bayes Algorithms

This study introduces a novel supervised learning algorithm with multiple enhancements designed specifically for MRI image analysis. After MRI images preprocessing, we use fuzzy clustering algorithm to segment them. This step helps to highlight the differences between pixel groups in different clusters. Once the dataset is enhanced, we extract the structural features from each image and estimate the probability density functions that represent each cluster. The prior probabilities of classified images are then determined, and the images are categorized using the Naïve Bayes method. The authors developed an efficient supervised learning algorithm for MRI image analysis building on these innovations. The proposed algorithm can be implemented on MATLAB. When tested on cancer imaging datasets, it consistently outperformed other classification methods like ResNet50 and VGG16, demonstrating its effectiveness and reliability.