3D Surface Texture Melanoma Diagnosis

Researchers at the University of the West of England have built a device that captures 3D skin surface texture images of dermatological lesions and aids in the diagnosis of melanoma. The device includes a camera and six LED light sources. A separate image is recorded with each of the LEDs illuminating the lesion. A computer then generates a 3D image of the lesions and extracts several features from it, such as asymmetry, border, color variation, and diameter. An artificial neural network and heuristic rules are used to identify malignant melanoma using these features. The system has an equal sensitivity to models based on 2D skin line pattern features, but with a greatly increased specificity. The researchers describe the technique in detail in the latest issue of the International Journal of Modelling, Identification and Control.
Study abstract: A computer assisted diagnosis system for malignant melanoma using 3D skin surface texture features and artificial neural network…
Project homepage: Application of photometric stereo in dermatology…