Brain Tumor Detection and Classifcation Using YOLO Algorithms

Numerous advances in healthcare technology have been made by applying image processing and computer vision for diagnosis. This is because MRI allows the radiologist to quickly and effectively identify even the smallest abnormalities in the brain. Building on the three popular object identification frameworks, YOLOv8, YOLOv9, and YOLOv11, This study employs YOLO algorithms to identify and categorize brain cancers in MRI images. It includes different types of brain tumors: pit form, also known as pituitary adenomas, brain form, gliomas, meningeal form, also known as meningiomas, and no tumor. The preprocessing step locates the tumor precisely by using advanced object detection features. This dataset cleaning and preprocessed by the opencv2 library has been used to assess how well the deep-learning YOLO models recognize and differentiate between brain tumors. The combined results of all YOLO models, we may infer from TABLE VIII YOLOv8 is the best at identifying the No Tumor and Pituitary classes, whereas YOLOv9 is the best at predicting the Meningioma class. The performance of YOLOv11 is the most accurate for the Meningioma and Pituitary classes when compared to YOLOv8. In general, YOLOv8 is more accurate at predicting stationary things than YOLOv9, and YOLOv11, which is more accurate at detecting small and shifting objects.