| Abstrakt: |
In this manuscript, a new computer-aided diagnosis method for melanoma is analyzed. This paper aims to improve some of the existing methods and develop new techniques to facilitate exact, prompt, and dependable computer-based diagnosis of melanoma. This approach makes contributions to various stages of a computer-aided diagnostic system of melanoma; namely, image segmentation or border detection, feature extraction, feature selection, and classification. This analysis system is estimated using a database of 32 cancerous and non-cancerous images. It is established to be operating satisfactorily with a detection accuracy of about 96%, Sensitivity of 96%, and Specificity of 100%. [ABSTRACT FROM AUTHOR] |
