Skin Melanoma, Actinic Keratosis, and Basal Cell Carcinoma Tumors Detections Versus Normal Tissues Based on A Deep Learning Classification Models

Skin tumor is one of the most common and dangerous malignancies affecting the human population. In particular, the 7 known cases of skin tumors are Melanoma, Melanocytic nevus, Basal cell carcinoma, Actinic keratosis / Bowen’s disease (intraepithelial carcinoma), Benign keratosis (solar lentigo/seborrheic keratosis/lichen planus-like keratosis), Dermatofibroma, and Vascular lesion. Various machine learning and deep learning models have been proposed over the last few years to help in early detection of skin tumors. Methods: We exploit the recent version of the YOLO series, YOLOv11, for the classification of benign and malignant cases of skin cancer. In addition, we evaluate other possible hybrid approaches utilizing EfficientNetV2M to efficiently tackle the problem and establish a profound understanding of the difference between YOLOv11 performance and that of other state-of-the-art architectures. Results: Our proposed deep learning model employing HAM10000 dataset achieves a YOLOv11.

testing accuracy of 92.8% and an EfficientNetV2M + NN testing accuracy score of 85.1%, respectively, demonstrating superior performance compared to previously established methods.