Toward automated detection of malignant melanoma

In vivo reflectance confocal microscopy shows promise for the early detection of malignant melanoma (MM). Two hallmarks of MM have been identified: the presence of pagetoid melanocytes in the epidermis and the breakdown of the dermal papillae. For detection of MM, these features must be identified q...

Full description

Bibliographic Details
Main Authors: Huang, Billy, Gareau, Daniel S.
Other Authors: Massachusetts Institute of Technology. Department of Biology
Format: Article
Language:en_US
Published: Society of Photo-optical Instrumentation Engineers 2010
Online Access:http://hdl.handle.net/1721.1/54722
Description
Summary:In vivo reflectance confocal microscopy shows promise for the early detection of malignant melanoma (MM). Two hallmarks of MM have been identified: the presence of pagetoid melanocytes in the epidermis and the breakdown of the dermal papillae. For detection of MM, these features must be identified qualitatively by the clinician and qualitatively through automated pattern recognition. A machine vision algorithm was developed for automated detection. The algorithm detected pagetoid melanocytes and breakdown of the dermal/epidermal junction in a pre-selected set of five MMs and five benign nevi for correct diagnosis.