Brain tumor detection is an important application of medical image analysis that can assist in identifying abnormal regions in brains cans.The present project focuses on developing an objectdetection system using YOLOv8 to localize brain-tumor regions in medical images. The project uses a YOLO-format dataset containing training, validation and test images with corresponding label files. The dataset was converted to zerobased YOLO class identifiers and a YOLOv8n model was configured for multi-class detection. The training workflow includes dataset organization, annotation verification, image preprocessing, augmentation, transfer learning, model training, validation and prediction. The working project records contain three target classes represented as Brain Tumor Type 0, Brain Tumor Type 1 and Brain Tumor Type 2. The available records show 2,451 training images, 306 validation images and 307 test images. The class distribution is not uniform across the splits, particularly in validation and test data, and this limitation must be considered when interpreting final performance.

Rakshith S ,Swetha S, Mathesh A, Sabarish M, Dhanya K R,"YOLOv8-Assisted Brain Tumor Detection" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(4): Page(655-671) July - August 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.