基于电阻抗扫描成像的乳腺癌检测方法

为了提高电阻抗成像在乳腺癌检测方面的性能,降低其假阳性率,提出了一种新的乳腺癌检测算子———异常能量指标(AEI)检测乳腺癌.通过对乳腺电阻抗成像进行建模并结合优化问题求解,提取乳腺癌病灶参数并进而构造AEI指标.与传统的基于临床医师视觉解释的图像检测方法相比,新方法可以获得更高的灵敏度、特异度和准确度.与基于参数的乳腺癌检测算法(HEDA算法和P算法)相比,AEI指标用于乳腺癌检测具有较高的灵敏度及特异度.临床数据实验表明,新方法可以有效地用于乳腺癌检测,具有较好的检测性能(灵敏度88.5%、特异度89.1%以及准确度88.9%).

Method for detecting breast cancer by eletrical impedance scanning

Abnormal energy indicator(AEI) calculated by complete parameters extraction algorithm(CPEA) is proposed to decrease the false positive rate of electrical impedance scanning.CPEA includes two parts: the modeling of the forward problem and the parameters extraction process.The parameters of the cancerous lesion within breast are extracted by CPEA which are then used to calculate the AEI for detection of breast cancer.Compared with the conventional detection method of image interpretation by experienced clinicians,using the AEI to detect breast cancer can achieve higher sensitivity,specificity and accuracy.In addition,employing the AEI can obtain excellent sensitivity and specificity compared with CAD(computer assisted diagnosis) algorithms,such as the HEDA(homologous electrical difference analysis) and the P algorithm.Experimental results suggest that AEI can be used in detecting breast cancer effectively,with a better detection performance(sensitivity as 88.5%,specificity as 89.1% and accuracy as 88.9%).