Vggnet for dermatological disease diagnosis with cloud-based image analysis

Dermatological disorders, and especially skin cancer, are a worldwide health issue. Precise and early diagnosis is critical in order to pursue efficient treatment, and machine learning, in this case Convolutional Neural Networks (CNNs), has tremendous potential in computerizing the diagnostic process. This research suggests a Visual Geometry Group Network (VGGNet) model for dermatological disorder diagnosis, specifically for skin disorders like melanoma, psoriasis, and eczema. The intended system utilizes cloud image processing for better diagnostic results and scalability. The process involved includes data retrieval, preprocessing (resizing, normalization, and augmentation), extraction of features based on VGGNet, classification with fully connected layers, and AWS-based storage in the cloud for data maintenance. The model demonstrates a noteworthy accuracy of 99%, even better than previous techniques like Hybrid GBDT+ALBERT+Firefly with 92% accuracy. Assessment of performance metric accuracy, precision, recall, and F1-score indicate that the suggested method performs better compared to others, with 96% precision, 97% recall, and 95.16% F1-score.