| dc.description.abstract | Basal cell carcinoma is a type of skin cancer that requires early detection as it can spread to surrounding tissues. The common symptoms of basal cell carcinoma include the appearance of small, pinkish bumps with a shiny surface on the skin., making the symptoms similar to actinic keratosis. Early detection and accurate diagnosis are crucial in the management of basal cell carcinoma. The diagnosis of basal cell carcinoma is performed using dermoscopy or histopathology methods, which require biopsy and can be time-consuming and costly. This study utilizes the EfficientNetV2 architecture, an improvement of the EfficientNet architecture, which has been proven to be efficient and effective in image processing, particularly in image classification. This architecture introduces several modifications to enhance model efficiency and accuracy. The model training uses the HAM 10000 2020 dataset, which contains 3000 data samples, divided into 2100 training data, 600 validation data, and 300 test data. The data undergoes preprocessing steps, including resizing, hair removal, segmentation, normalization, and augmentation. After preprocessing, the data is fed into the EfficientNetV2 architecture for feature extraction and model training, which uses 40 epochs and a batch size of 32. The results of using the EfficientNetV2 architecture show an accuracy of 94%. Based on this accuracy result, it can be concluded that the system performs very well in identifying basal cell carcinoma through dermoscopic images. | en_US |
