YOLOv1 to YOLOv10: A Comprehensive Review of YOLO Variants and Their Application in Medical Image Detection
Ahealisi Yeerjiang, Zongyu Wang, Xiangtong Huang, Jing Zhang, Qi Chen, Yucheng Qin, Jia He
Abstract
The rapid evolution of computer vision has elevated object detection to a central task within the field. In medicine, automated lesion detection has the potential to greatly improve diagnostic efficiency for clinicians. The extraordinary success of deep learning in computer vision has motivated researchers globally to apply these advancements to medical image analysis. Deep learning techniques have demonstrated superior performance in medical image classification, detection, segmentation, registration, and retrieval compared to traditional methods. Among these, the YOLO (You Only Look Once) series of algorithms stands out for their exceptional speed and accuracy, making them a popular choice for medical image detection. This paper presents the underlying principles and structure of the classic YOLO algorithms, reviews their current applications in medical image detection, addresses existing challenges, and explores future directions for the application of YOLO in this domain.