This work presents a simplified approach for the early detection of breast cancer using thermographic images, showing that several thermophysical properties of the bio-thermal problem does not need to be previously known to locate the geometric center of tumors. A 3D hemispheric breast model composed of different layers (muscle, gland, fat and skin) was constructed to evaluate the thermal behavior on the skin surface from numerical simulations using commercial software COMSOL. The effects of changes in depth, size, metabolism, blood perfusion and thermal conductivity of the tumor at surface temperatures were systematically analyzed to provide important information and guidelines for future medical diagnoses. Variations in blood perfusions and thermal conductivities of healthy tissue layers were also evaluated. It has been found that changing the size, metabolism, blood perfusion and thermal conductivity of a centralized tumor in the same coordinate does not modify the profiles of normalized temperature variations on the breast skin surface. Regarding the properties of healthy tissue, if a specific region of the breast surface is taken, there is the possibility that the normalized temperature profiles also do not depend on these properties. Thus, knowing that one of the main limitations in the estimation of tumors from thermographic images is related to the difficulty of previously knowing the thermophysical properties of human tissues, the results obtained in this study provide valuable simplifications for the early diagnosis of breast cancer using infrared thermography.
Read full abstract