Enhanced 3D curvature pattern and melanoma diagnosis

Warr, Robert; Zhou, Yu; Smith, Melvyn; Smith, Lyndon; Farooq, Abdul

Abstract

This article describes an enhanced curvature pattern based melanoma diagnosis system using convolution techniques and ensemble classifiers. We extract the 3. D data of melanoma with a photometric stereo device first. Then differential forms of the melanoma surface can be extracted with the convolution method proposed. After extracting 3. D based differential forms, statistical moments of enhanced principal curvatures of skin surfaces are calculated to describe the geometrical texture patterns. Finally, ensemble classifiers are constructed whose optimal mean sensitivity and specificity can reach 89.24 percent and 87.62 percent respectively. Comparisons with skin tilt/slant pattern based 3. D shape characterization method and 2. D methods like color variation and border irregularity are also included. © 2010 Elsevier Ltd.

• Journal Article Type: Article
• Publication Date: Mar 1, 2011
• Deposit Date: Nov 27, 2012
• Journal: Computerized Medical Imaging and Graphics
• Print ISSN: 0895-6111
• Electronic ISSN: 1879-0771
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 35
• Issue: 2
• Pages: 155-165
• DOI: https://doi.org/10.1016/j.compmedimag.2010.10.004
• Keywords: Shape analysis, melanoma, convolution, curvature, ensemble classifier
• Public URL: https://uwe-repository.worktribe.com/output/974280
• Publisher URL: http://dx.doi.org/10.1016/j.compmedimag.2010.10.004
• Contract Date: Apr 12, 2016