Discrimination of Benign and Malignant Suspicious BreastTumors Based on Semi-Quantitative DCE-MRI ParametersEmploying Support Vector Machine

Dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) is aneffective tool for detection and characterization of breast lesions. Qualitative assessmentof suspicious breast DCE-MRI is problematic and operator dependent. The purposeof this study is to evaluate diagnostic efficacy of the representative characteristicparameters, extracted from kinetic curves of DCE-MRI, for discrimination betweenbenign from malignant suspicious breast tumors. Pre-operative DCE-MR images of twenty-six histopathological approvedbreast lesions were analyzed. The images were reviewed by an expert radiologist andthe regions of interests (ROI)s were selected on the most solid part of the lesion. Semiquantitativekinetic parameters, namely: maximum signal enhancement (SImax), initialarea under the curve (IAUC60), time to peak (TTP), wash in rate (WIR), wash out rate(WOR) and signal enhancement ratio (SER), were calculated within each ROI. Meanvalues of the calculated features among benign and malignant groups were comparedusing student’s t-test. Finally, a classification was performed employing supportvector machines (SVM) using each of the parameters and their combinations in orderto investigate the efficacy of the parameters in distinguishing between benign frommalignant tumors. The performance of the classification procedure employing the combinationof semi-quantitative features with (p-value<0.001) was evaluated by means of severalmeasures, including accuracy, sensitivity, specificity, positive predictive value andnegative predictive value which returned amounts of 97.5%, 96.49%, 100%, 100%and 95.61% respectively. In conclusion, semi-quantitative analysis of the characteristic kineticcurves of suspicious breast lesions derived from SVM classifier