Classification of Liver Tumors Based on YOLOv8s-cls

As liver cancer incidence and mortality rates continue to increase, the early diagnosis and treatment of liver tumors are essential for effective management and accurate classification of the disease. However, distinguishing between early and late-stage liver tumors remains challenging. To tackle the issues of low accuracy and slow classification speed in liver tumor identification, this study introduces the YOLOv8s-cls model, which is both lightweight and fast, for classifying benign and malignant tumors. Evaluation results indicate that this model demonstrates impressive classification capabilities, achieving an average accuracy of 94.3%, a precision of 95.4%, a recall of 96.1%, and an F1-score of 94.6%. When compared to ResNet-50, MobileNet V3, and ShuffleNet V2, this model excels in both accuracy and speed. Thus, it presents a dependable approach for classifying benign and malignant liver tumors.