Computer-aided diagnosis of psoriasis classification using deep learning techniques

Psoriasis is a long-term autoimmune skin disorder that impacts millions of individuals globally. Early diagnosis of psoriasis enables timely intervention, helping to prevent the progression of the condition and manage symptoms more efficiently. In this research, proposed an algorithm for Psoriasis Detection System using CNN-VGG16 algorithm, AlexNet and SqueezeNet to automate the identification of different psoriatic lesions in skin images. The CNN architecture is built to automatically extract distinguishing features from input images, allowing the system to differentiate between normal and psoriatic skin patterns. The model undergoes rigorous training and validation processes to optimize its performance. The Proposed AlexNet and SqueezNet both demonstrated good results with accuracies of 94.4% and 96.3%, respectively, but the CNN-VGG16 surpassed all other models with an accuracy of 97.22%. The outcomes of this research contribute to the advancement of automated psoriasis detection systems, offering a reliable and efficient tool for dermatologists and healthcare professionals.