Noise Estimation in Infrared Image Sequences: A Tool for the Quantitative Evaluation of the Effectiveness of Registration Algorithms

Dynamic infrared imaging has been proposed in literature as an adjunctive technique to mammography in breast cancer diagnosis. It is based on the acquisition of hundreds of consecutive thermal images with a frame rate ranging from 50 to 200 frames/s, followed by the harmonic analysis of temperature...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on biomedical engineering Vol. 55; no. 7; pp. 1917 - 1920
Main Authors Agostini, Valentina, Delsanto, Silvia, Knaflitz, Marco, Molinari, Filippo
Format Journal Article
LanguageEnglish
Published United States IEEE 01.07.2008
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Algorithms
Blood
Body Temperature
Breast cancer
Breast cancer detection
dynamic infrared imaging (DIRI)
Fluctuations
Harmonic analysis
Humans
Image Interpretation, Computer-Assisted - methods
image registration
Image sequences
Infrared imaging
Infrared Rays
Mammography
Pixel
Reproducibility of Results
Sensitivity and Specificity
Skin
Skin Physiological Phenomena
SNR estimation
Studies
Subtraction Technique
Temperature distribution
thermal image sequence
Time series
Breast cancer detection
dynamic infrared imaging
signal-to-noise ratio estimation
thermal image sequence
image registration
Online AccessGet full text

Cover

More Information
Summary:Dynamic infrared imaging has been proposed in literature as an adjunctive technique to mammography in breast cancer diagnosis. It is based on the acquisition of hundreds of consecutive thermal images with a frame rate ranging from 50 to 200 frames/s, followed by the harmonic analysis of temperature time series at each image pixel. However, the temperature fluctuation due to blood perfusion, which is the signal of interest, is small compared to the signal fluctuation due to subject movements. Hence, before extracting the time series describing temperature fluctuations, it is fundamental to realign the thermal images to attenuate motion artifacts. In this paper, we describe a method for the quantitative evaluation of any kind of feature-based registration algorithm on thermal image sequences, provided that an estimation of local velocities of reference points on the skin is available. As an example of evaluation of a registration algorithm, we report the evaluation of the SNR improvement obtained by applying a nonrigid piecewise linear algorithm.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ObjectType-Article-1
ObjectType-Feature-2
ISSN:0018-9294
1558-2531
DOI:10.1109/TBME.2008.919842