Advanced infrared image processing for breast cancer risk assessment

Abstract

The following report describes image processing techniques applied to high resolution breast thermograms as used to predict breast cancer. The goal is to observe asymmetry in the heat pattern due to temperature differences and/or the areas of observed vesicular structure and other "hot spots". Temperature and hot spot area differences are computed between the patient's left and right breasts and structurally matched breast quadrants. The approach is designed to generate objective measures for determining the patient's cancer risk.