Boosting Brain Tumor Diagnosis with Smart AI

Brain tumors are super serious and need quick, accurate diagnosis for the best treatment. Old ways for classifying these tumors take time and can be wrong. Let's start with collecting brain images from a special dataset. These images are then cleaned up to remove noise using a special method called trust-based distributed set-membership filtering (TDSF). After this, we use quaternion offset linear canonical transform (QOLCT) to grab special features from the images, like Grayscale stats and textures. Next, these features are fed into a clever AI called Semantic-Preserved Generative Adversarial Network (SPGAN). This AI helps classify tumors into three types: Glioma, Meningioma, and Pituitary. To make the AI even smarter, we use Hunger Games Search Optimization (HGSO) to tweak its settings. This new method, BTC-SPGAN-HGSO, does really well. It can tell the difference between tumor types with high accuracy: 99. 72% for Glioma, 99. 65% for Meningioma, and 99. 52% for Pituitary. Plus, it has very low error rates. This new approach helps doctors and neurologists make the right calls about diagnosing brain tumors faster and better than before.