DIFFERENTIATION OF BENIGN AND MALIGNANT BREAST-TUMORS BY LOGISTIC-REGRESSION AND A CLASSIFICATION TREE USING DOPPLER FLOW SIGNALS

In breast examinations with Doppler, an increased flow is found in mal ignant tumors. With the relatively new color Doppler, we measured diff erent flow values in 133 cancer patients and in 325 women with benign disease. These measurements were used to develop diagnostic rules. For the highly correlated flow values, we used a stepwise procedure to !; elect a final logistic regression model and a tree-based approach, whi ch is a different way of modeling. With both approaches we developed s imple diagnostic rules of which the sensitivity and the specificity ex ceeded 90%. There are no differences between the two approaches concer ning discriminative ability. As complex statistical modeling leads to an overoptimism in the assessment of the error rates, we applied sensi tivity analysis, investigated the stability of the selected logistic r egression model, and estimated the magnitude of the overoptimism of th e diagnostic rules with resampling methods. The results indicate that the estimates of sensitivity and specificity are probably close to rea listic values for a clinical setting.