Masters Theses

Abstract

Melanoma is a form of a skin cancer, which can be deadly if not treated in the early stages. It is believed that color plays an important role in the diagnosis of malignant melanoma and that there are “melanoma colors” lying in the color space. The main objective of this research is to determine those “melanoma colors” and to find the locations of melanoma-colored pixels throughout the tumor. A color map for melanoma pixels is calculated using a training set. With this color map, different images are scored and for those images the percentage of melanoma-colored pixels are calculated. This color map is also used to derive spatial domain information for melanoma-colored pixels. This information is used to calculate two features called “clustering” and “asymmetry”. Before this work was done, it was believed that the melanoma-colored pixels were mostly clustered in a tumor. The results verified this assumption. The results also indicated that this clustering occurred more intensely in melanomas than in benign lesions. The “asymmetry” feature did not suggest a reliable way to diagnose melanoma. This research dealt with both clinical and dermatoscopy images. Dermatoscopy images were diagnosed better than the clinical images

Advisor(s)

Moss, Randy Hays, 1953-

Committee Member(s)

Stanley, R. Joe
Hering, Roger H.

Department(s)

Electrical and Computer Engineering

Degree Name

M.S. in Electrical Engineering

Publisher

University of Missouri--Rolla

Publication Date

Fall 2000

Pagination

vi, 50 pages

Note about bibliography

Includes bibliographical references (page 49).

Rights

© 2000 Yunus Emre Faziloglu, All rights reserved.

Document Type

Thesis - Restricted Access

File Type

text

Language

English

Thesis Number

T 7966

Print OCLC #

48797323

Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.

Share

 
COinS