Utilizing YOLOv5x for the Detection and Classification of Brain Tumors
M. Kumar, Urmila Pilania, Tanisha Bhayana, Stuti Thakur
Abstract
Development in image processing and Deep Learning (DL) techniques have many applications in the medical field. In this work, YOLOv5x (You Only Look Once) has been used for brain tumor detection and classification in Magnetic Resonance Imaging (MRI) images. YOLOv5x is well recognized for its accuracy and speed in real-time object detection. The proposed work comprises training the YOLOv5x model on the Brats and Roboflow dataset of brain MRI images. It is trained to identify and classify types of brain tumors. With the proper training process, the accuracy and speed of detecting irregularities in several tumor cases can be improved. Performance metrics like precision, recall, mAP (Mean Average Precision), and F1-Score are utilized to measure the robustness and reliability of the model. For validation of the proposed work, simulation results are computed using two datasets. The results of the work hold significant potential for timely recognition and diagnosis of brain tumors.