A machine learning method for identifying morphological patterns in reflectance confocal microscopy mosaics of melanocytic skin lesions in-vivo

Kivanc Kose; Christi Alessi-Fox; Melissa Gill; Jennifer G. Dy; Dana H. Brooks; Milind Rajadhyaksha

doi:10.1117/12.2212978

29 February 2016 A machine learning method for identifying morphological patterns in reflectance confocal microscopy mosaics of melanocytic skin lesions in-vivo

Kivanc Kose, Christi Alessi-Fox, Melissa Gill, Jennifer G. Dy, Dana H. Brooks, Milind Rajadhyaksha

Author Affiliations +

Proceedings Volume 9689, Photonic Therapeutics and Diagnostics XII; 968908 (2016) https://doi.org/10.1117/12.2212978
Event: SPIE BiOS, 2016, San Francisco, California, United States

Abstract

We present a machine learning algorithm that can imitate the clinicians qualitative and visual process of analyzing reflectance confocal microscopy (RCM) mosaics at the dermal epidermal junction (DEJ) of skin. We divide the mosaics into localized areas of processing, and capture the textural appearance of each area using dense Speeded Up Robust Feature (SURF). Using these features, we train a support vector machine (SVM) classifier that can distinguish between meshwork, ring, clod, aspecific and background patterns in benign conditions and melanomas. Preliminary results on 20 RCM mosaics labeled by expert readers show classification with 55 − 81% sensitivity and 81 − 89% specificity in distinguishing these patterns.

Conference Presentation

Citation Download Citation

Kivanc Kose, Christi Alessi-Fox, Melissa Gill, Jennifer G. Dy, Dana H. Brooks, and Milind Rajadhyaksha "A machine learning method for identifying morphological patterns in reflectance confocal microscopy mosaics of melanocytic skin lesions in-vivo", Proc. SPIE 9689, Photonic Therapeutics and Diagnostics XII, 968908 (29 February 2016); https://doi.org/10.1117/12.2212978

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available