文章摘要

文章信息/Info

Title:: Investigation of Multivariable Estimation in Breast Tumor Diagnosis by Infrared Thermography

Author(s):: HUANG Jian; LU Mei; LI Bohan; ZHANG Tao; School of Energy and Power Engineering, University of Shanghai for Science and Technology，Shanghai 200093, China

Keywords:: Breast tumor; Infrared diagnosis; Multivariable inverse heat conduction problem; Multi fitness function; Particle swarm optimization algorithm

摘要:: 根据红外热像仪测得的体表温度来获得肿瘤相关信息，可以抽象为一个含有内热源的导热反问题。针对同时反演肿瘤位置、肿瘤代谢热强度及乳腺导热系数等多参数反问题的特点，本研究提出并建立了与反演参数相关的辅助目标函数，同时对求解所采用的粒子群优化算法中的更新方式进行了相应的改进。数值验证结果表明：在反演多参数反问题中采用多目标函数能提高反演精度，加快收敛速度。研究结果为根据乳腺红外热像图量化诊断体内肿瘤提供了理论依据。

Abstract:: Medical infrared thermography is a non-contact, non-invasive imaging method that applied in the area of breast cancer detection. With thermogram, parameters of tumor inside can be estimated, and an inverse heat conduction model with inner heat source was obtained from it. Embark upon the situation that inverse heat conduction problem has multi variables, tumor position, metabolic heat generation rate and thermal conductivity, two more assistant fitness function were established base on how heat source position and conductivity influence the surface temperature, and Particle Swarm Optimization was modified to solve the multivariable inverse problem. Numerical verification showed that the method was suitable for breast tumor properties estimation, and the modified PSO with multi fitness functions improved the accuracy and shortened the solution time. The results provide theoretical basis for quantitative diagnosis of breast tumor by infrared thermal images.