Application of AutoML, YOLOv5, and YOLOv8 in brain MRI tumor image recognition

With the rapid development of machine learning, particularly deep learning, the automatic recognition and classification of brain tumors have become a significant focus of research in medical image analysis. However, due to the complex morphology and indistinct boundaries of brain tumors, traditional manual annotation methods are time-consuming and prone to errors, highlighting the need for more intelligent diagnostic tools. This study compares the effectiveness of applying AutoML and YOLO series models (YOLOv5 and YOLOv8) in brain tumor recognition, aiming to evaluate the performance of each model and provide insights for improving diagnostic accuracy and efficiency. The results indicate that YOLOv8 outperforms the other models in terms of precision, recall, and F1-score, making it highly suitable for high-precision applications. YOLOv5, with its lightweight design, offers advantages in inference speed. While EasyDL performs slightly lower in certain metrics, its high precision and ease of use make it a competitive option in specific scenarios.The main contribution of this paper lies in not only verifying the high feasibility of the emerging AutoML technology with weaker requirements on expert knowledge in the field of MRI image recognition for brain tumors, but also providing a detailed comparative analysis that can serve as a reference for medical practitioners to flexibly choose suitable model algorithms based on their cost, demand, etc. conditions.