This paper reports on a series of new trends in the area of breast thermography that aim to improve accuracy in the early detection of breast cancer. The introductory section presents information about the advantages offered by thermography, mainly due to its noninvasiveness and lower cost. This is accompanied by a comparison between thermography and more traditional mammography in terms of performance indicators such as sensitivity and specificity. Since the ’80s, the techniques for computerized image processing have become more and more popular, enabling physicians to enhance medical diagnosis with computer-assisted detection/diagnosis (CAD).
The authors present some historical details concerning the basic tools used to design and implement thermographic CAD systems, such as artificial neural networks and fuzzy systems. They discuss their conclusions about the performance of CAD in breast thermography, based on statistical evidence, and explore the advantages of using hybrid techniques in designing more accurate thermographic CAD systems.
The list of bibliographic references contains several standard titles. Unfortunately, some of the cited references are missing (indicated by a question mark in the text).
This paper will mainly be of interest to readers who would like a general introduction to the methodologies currently used for CAD in breast thermography.