##plugins.themes.bootstrap3.article.main##

Haewon Byeon

Abstract

This study explores the application of the YOLO v10 model for the detection and classification of brain tumors in CT images. YOLO, known for its real-time object detection capabilities, offers a promising approach to addressing the challenges of medical imaging. The research utilizes the Brain Tumor Dataset from Kaggle, incorporating 437 negative and 488 positive images for training, with additional datasets for validation. The YOLO v10 model demonstrated superior performance compared to traditional models like AlexNet, VGG16, ResNet101V2, and MobileNetV3-Large. It achieved a precision of 0.920, recall of 0.890, F1-score of 0.900, and accuracy of 0.910. These results highlight its effectiveness in accurately identifying and classifying tumors, offering significant potential for clinical applications. The model's architecture allows for efficient processing of high-resolution CT scans, adapting well to varied tumor sizes and shapes. The study also discusses the challenges and future directions for improving computational efficiency and generalization capability in diverse datasets. The promising findings suggest that YOLO v10 can be a powerful tool in medical diagnostics, enhancing the accuracy and speed of tumor detection and contributing to better patient outcomes. This research sets a foundation for further exploration and development of YOLO-based models in healthcare.

##plugins.themes.bootstrap3.article.details##