Dermoscopy analysis of RGB-images based on comparative features

Oleg O. Myakinin; Valery P. Zakharov; Ivan A. Bratchenko; Dmitry N. Artemyev; Evgeny Y. Neretin; Sergey V. Kozlov

doi:10.1117/12.2188165

22 September 2015 Dermoscopy analysis of RGB-images based on comparative features

Oleg O. Myakinin, Valery P. Zakharov, Ivan A. Bratchenko, Dmitry N. Artemyev, Evgeny Y. Neretin, Sergey V. Kozlov

Proceedings Volume 9599, Applications of Digital Image Processing XXXVIII; 95992B (2015) https://doi.org/10.1117/12.2188165
Event: SPIE Optical Engineering + Applications, 2015, San Diego, California, United States

Abstract

In this paper, we propose an algorithm for color and texture analysis for dermoscopic images of human skin based on Haar wavelets, Local Binary Patterns (LBP) and Histogram Analysis. This approach is a modification of «7-point checklist» clinical method. Thus, that is an “absolute” diagnostic method because one is using only features extracted from tumor’s ROI (Region of Interest), which can be selected manually and/or using a special algorithm. We propose additional features extracted from the same image for comparative analysis of tumor and healthy skin. We used Euclidean distance, Cosine similarity, and Tanimoto coefficient as comparison metrics between color and texture features extracted from tumor’s and healthy skin’s ROI separately. A classifier for separating melanoma images from other tumors has been built by SVM (Support Vector Machine) algorithm. Classification’s errors with and without comparative features between skin and tumor have been analyzed. Significant increase of recognition quality with comparative features has been demonstrated. Moreover, we analyzed two modes (manual and automatic) for ROI selecting on tumor and healthy skin areas. We have reached 91% of sensitivity using comparative features in contrast with 77% of sensitivity using the only “absolute” method. The specificity was the invariable (94%) in both cases.

Citation Download Citation

Oleg O. Myakinin, Valery P. Zakharov, Ivan A. Bratchenko, Dmitry N. Artemyev, Evgeny Y. Neretin, and Sergey V. Kozlov "Dermoscopy analysis of RGB-images based on comparative features", Proc. SPIE 9599, Applications of Digital Image Processing XXXVIII, 95992B (22 September 2015); https://doi.org/10.1117/12.2188165

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