Texture in Skin Images: Comparison of Three Methods to Determine Smoothness

Abstract

Smooth texture, a critical feature in skin tumor diagnosis, is analyzed using three texture measurement methods. A dermatologist classified 1290 small blocks within 42 tumor images as smooth, partially smooth, or nonsmooth. Texture discriminatory power of three methods were compared: the neighboring gray-level dependence matrix (NGLDM) method of Sun and Wee, the circular symmetric autoregressive random field model of Kashyap and Khotanzad, and a new peak-variance method. The texture analysis method that allows best prediction of smoothness for our tumor domain is the NGLDM method, affording 98% correct prediction of a smooth block with 21% false positives. We discuss applicability of texture analysis to dermatology.

Department(s)

Chemistry

Second Department

Electrical and Computer Engineering

Sponsor(s)

National Science Foundation (U.S.). Small Business Research Innovation Program

Comments

This research was supported by the National Science Foundation small business innovation research grant ISI 8521284.

Keywords and Phrases

Biological Materials - Textures; Biomedical Engineering - Diagnosis; Biomedical Engineering - Oncology; Image Processing - Image Analysis; Surfaces - Roughness Measurement; Circular Symmetric Autoregressive Random Field Model; Neighboring Gray Level Dependence Matrix (NGLDM); New Peak Variance Method; Skin Tumor Diagnosis; Texture Analysis; Biological Materials; Article; Cancer Classification; Cancer Diagnosis; Computer Graphics; Dermatology; Diagnostic Imaging; Digital Computer; Image Analysis; Medical Photography; Medical Technology; Melanoma; Priority Journal; Skin Cancer; Algorithms; Comparative Study; Diagnosis, Computer-Assisted; Human; Image Processing, Computer-Assisted; Models, Biological; Palpation; Photography; Predictive Value Of Tests; Signal Processing, Computer-Assisted; Skin Neoplasms; Support, U.S. Gov't, Non-P.H.S.; Computer Diagnosis; Skin Imaging; Smooth

International Standard Serial Number (ISSN)

0895-6111; 1879-0771

Document Type

Article - Journal

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 1992 Elsevier, All rights reserved.

Publication Date

01 May 1992

PubMed ID

1623493

Share

 
COinS