Automated framework for detection and classification of breast abnormality using breast thermograms

Abstract

Among all the cancers that affect women breast cancer is the most newlineprevalent and widely regarded as invasive cancer worldwide Breast cancerhas a great probability of being cured if diagnosed early As a result a properscreening technique leads to an earlier disease diagnosis Mammography haslong been regarded as the most effective screening technology in the world However it has drawbacks such as an expensive and painful imaging technique significant radiation exposure and low screening accuracy inwomen under the age of 40 and women with dense breastsOn the other hand breast thermography is noninvasive noncontactand non-radiative making it feasible to screen women of any agegroup and reduce the risk of damaging radiation to pregnant and lactatingwomen Hotspots on a thermogram reflect abnormal regions in the breast that are warmer than the surrounding normal regions A rise in local temperatureis thought to be one of the first signs of a growing malignancy The surfacetemperature differences in breast thermograms are recorded in the form ofintensity variations that occur at the initial stages of cancer which aids in theearly diagnosis of breast cancer Interpretations of breast thermograms are subjective to the professionals knowledge ambient factors affect imagequality so despite its several advantages it is not widely employed as a cancer screening tool newline newline