Journal of Emerging Trends in Engineering and Applied Sciences: Research Areas

Article Title: Automatic Detection of Early-Stage Malignant Melanoma from Skin Lesion Images
by Nwawudu Sixtus Ezenwa, and Prof. Amine Nalt-Ali

Abstract:
The automatic detection of early stage malignant melanoma will give potential hope for diagnosis for melanoma skin cancer. This paper will also find a new method to analyse skin lesion images and this can enhance the dermatology department. Melanoma is the most dangerous form of skin cancer, accurate diagnosis is required to control the disease. So, the chances of false detection due to human error are high in a large population to be screened due to high workload on dermatologist and the objectivity in the interpretation of the screening which in turn can lead to fatal condition. The accurate and timely diagnosis of melanoma infection is essential to control and cure the disease. This study aims to explore the possibility of computerised diagnosis of early-stage malignant melanoma and to develop a novel image processing algorithm to reliably detect the presence melanoma from a sample skin image. Some image processing algorithms to automate the diagnosis of early stage melanoma on the skin are developed. This study curbs the human error while detecting the presence of early stage melanoma on the skin using image processing and automation. We achieved this goal using Image Segmentation and Morphological features descriptors. We also built the system in a robust manner so that it is unaffected by the exceptional conditions and achieved high percentage of sensitivity, specificity, positive prediction and negative prediction results.

Keywords: dermatoscopy automatic image processing; biomedical image processing; dermatologist; segmentation; features descriptors.

Download full paper