Year: 2025
Type: article
Abstract: Skin cancer is a serious problem that is frequently ignored. Sometimes, when a doctor does a manual examination, the human eye cannot accurately identify illnesses from imaging data. Deep learning techniques are increasingly being used in today's world to solve problems in our daily lives. Thus, we use deep neural network methods to create an automated computerized system for identifying skin conditions. We employed a number of neural network algorithms in the suggested model, analyzed their results, and determined which algorithm performed the best in terms of accuracy in detecting the five main skin conditions. CNN, and we have developed a new model to achieve an accuracy of about 80% by utilizing the Keras Sequential API. Later, to improve accuracy and for comparison, we have employed pre-trained data-based architectures. These deep learning models consist of DENSENET201 andinception-v3. The ResNet architecture achieves the maximum accuracy of 97% among the algorithms employed in the suggested models.
DOI: 10.71146/kjmr407
Authors: Aiman Ali Batool, Hasnaib Khan, Saima Ali Batool, Mohsin Ali Tariq, Naeem Aslam, Haseeb Ur Rehman
Language: English
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W4410008372