Title: Differentiation of breast abnormalities in infrared images using Riesz and quaternion Hilbert transform based features

Authors: S. Prabha; C.M. Sujatha; S.S. Suganthi

Addresses: Department of Electronics and Communication Engineering, College of Engineering Guindy, Anna University, Chennai, Tamil Nadu, India ' Department of Electronics and Communication Engineering, College of Engineering Guindy, Anna University, Chennai, Tamil Nadu, India ' Non Invasive Imaging and Diagnostics Laboratory, Biomedical Engineering Group, Department of Applied Mechanics, Indian Institute of Technology Madras, Chennai, Tamil Nadu, India

Abstract: In this paper, an attempt is made to characterise the normal and abnormal breast tissues in infrared images using Riesz and quaternion Hilbert transform based features. Left and right breast regions are separated using midpoint of inframammary folds. Riesz and quaternion based features are extracted from these separated breast tissues. An index based on the extracted features is calculated and proposed for breast tissue analysis. The results show that the features extracted using Riesz and quaternion Hilbert transform are able to discriminate normal and pathological conditions of breast tissues. The extracted features could capture the thermal variations among various abnormalities that are introduced due to varied metabolic conditions. Also, the index calculated using quaternion features is found to be statistically significant in distinguishing normal and carcinoma tissues. Hence, it appears that the proposed index using quaternion features can be useful for efficient diagnosis of breast pathologies.

Keywords: breast infrared images; quaternion Hilbert transform; Riesz transform; feature index; breast abnormalities; breast cancer; feature extraction; thermal images; carcinoma tissues.

DOI: 10.1504/IJBET.2015.072995

International Journal of Biomedical Engineering and Technology, 2015 Vol.19 No.3, pp.255 - 265

Received: 20 Mar 2015
Accepted: 04 Jun 2015

Published online: 11 Nov 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article