Computer-assisted diagnosis of melanoma.

The computer-assisted diagnosis of melanoma is an exciting area of research where imaging techniques are combined with diagnostic algorithms in an attempt to improve detection and outcomes for patients with skin lesions suspicious for malignancy. Once an image has been acquired, it undergoes a processing pathway which includes preprocessing, enhancement, segmentation, feature extraction, feature selection, change detection, and ultimately classification. Practicality for everyday clinical use remains a vital question. A successful model must obtain results that are on par or outperform experienced dermatologists, keep costs at a minimum, be user-friendly, and be time efficient with high sensitivity and specificity.