IMECE2003-43825 NUMERICAL SIMULATION OF ENHANCED SKIN THERMAL SIGNATURE OF FEMALE BREAST TUMOR SUBJECTED TO FORCED CONVECTION

ABSTRACT Early detection is considered to be the best defense against breast cancer and imaging plays a very important role in screening and in the diagnosis of symptomatic women. Infrared thermal imaging of skin temperature changes caused by a malignant tumor in breast is a rapidly developing detection modality with potential for functional detection. Knowledge and control of environmental factors which affect the skin temperature can reduce misinterpretations and false diagnosis associated with infrared imaging. A bio heat transfer based numerical model was utilized to study the energy balance in healthy and malignant breasts subjected to low velocity forced convection in a wind tunnel. Existing estimates of metabolic heating rates and previous measurements of temperature distributions along the radial direction in a region intersecting a known tumor and a comparable region in the healthy breast of the same patient were used to estimate the blood perfusion rates for the tumor. A simplified structural and thermal model was used for representing the changes within and around the tumor. Steady state temperature distributions on the skin surface of the breasts were obtained by numerically solving the conjugate heat transfer problem. Parametric studies on the influences of the airflow on the skin thermal expression of tumors were performed. It was found that the presence of tumor may not be clearly shown due to the irregularity of the skin temperature distribution induced by the flow field. Image processing techniques could be employed to eliminate the effects of the flow field and thermal noise and significantly improve the thermal signature of the tumor on the skin surface