Derm Detect; an AI Powered Application for Skin Cancer Detection

Skin cancer is one of the most rampant forms of cancer, with millions of people been diagnosed yearly. Early detection of cancer, makes it possible for patients to receive required treatment and adequately manage the illness, with few making full recovery. However, traditional methods of detecting cancer, such as visual examination by dermatologists are time-consuming and subjective. This often leads to errors resulting in misdiagnosis or false negatives, allowing malignant cells to go undetected until metastasis occurs making it too late for treatment. The evolution of image processing and machine learning techniques has opened the way for automated diagnosis systems. This thesis presents an AI powered web application called “DERM DETECT” through which skin lesions can be detected automatically. This was achieved by integrating computer vision and deep learning technology. The proposed application demonstrated good diagnostic performance with an accuracy of 9%. Derm Detect will revolutionize existing clinical methods of detecting skin cancer and promote routine self-screening examinations by individuals, while advising patients to seek further medical attention if needed. Furthermore, the application will assist dermatologists easily evaluate photos of various skin problems and accurately detect any skin tumor, thus decreasing human error, improving diagnostic reliability and accuracy, The implementation of the proposed apps for use in hospitals and for self-examination by individuals, will promote early identification and treatment of patients at risk, ultimately reducing the incidence of skin cancer-related morbidity and death thus saving lives.