Computerized characterization of breast masses on three‐dimensional ultrasound volumes

Abstract

We are developing computer vision techniques for the characterization of breast masses as malignant or benign on radiologic examinations. In this study, we investigated the computerized characterization of breast masses on three-dimensional (3-D) ultrasound (US) volumetric images. We developed 2-D and 3-D active contour models for automated segmentation of the mass volumes. The effect of the initialization method of the active contour on the robustness of the iterative segmentation method was studied by varying the contour used for its initialization. For a given segmentation, texture and morphological features were automatically extracted from the segmented masses and their margins. Stepwise discriminant analysis with the leave-one-out method was used to select effective features for the classification task and to combine these features into a malignancy score. The classification accuracy was evaluated using the area Az under the receiver operating characteristic (ROC) curve, as well as the partial area index Az(0.9), defined as the relative area under the ROC curve above a sensitivity threshold of 0.9. For the purpose of comparison with the computer classifier, four experienced breast radiologists provided malignancy ratings for the 3-D US masses. Our dataset consisted of 3-D US volumes of 102 biopsied masses (46 benign, 56 malignant). The classifiers based on 2-D and 3-D segmentation methods achieved test Az values of 0.87+/-0.03 and 0.92+/-0.03, respectively. The difference in the Az values of the two computer classifiers did not achieve statistical significance. The Az values of the four radiologists ranged between 0.84 and 0.92. The difference between the computer's Az value and that of any of the four radiologists did not achieve statistical significance either. However, the computer's Az(0.9) value was significantly higher than that of three of the four radiologists. Our results indicate that an automated and effective computer classifier can be designed for differentiating malignant and benign breast masses on 3-D US volumes. The accuracy of the classifier designed in this study was similar to that of experienced breast radiologists.

Malignant and Benign Breast Masses on 3D US Volumetric Images: Effect of Computer-aided Diagnosis on Radiologist Accuracy

“…We have developed an automated computer classifier for differentiation of malignant and benign breast masses on threedimensional (3D) US volumetric images (15). Thus, the purpose of our study was to retrospectively investigate the effect of using the computer classifier we developed on radiologists' sensitivity and specificity for discriminating malignant masses from benign masses on 3D volumetric US images, with histologic analysis serving as the reference standard.…”

“…The first step of computerized analysis (15) involved extraction of the mass boundaries in the 3D volumetric data set (ie, mass segmentation). Automated segmentation of breast masses on US images is a difficult task because of image speckles, posterior shadowing, and variations of the gray level both within the mass and within the normal breast tissue.…”

Reliable Evaluation of Performance Level for Computer-Aided Diagnostic Scheme

“…Improper use of these methods can lead to significant biases in the estimated performance levels. We reviewed all research articles on the development and evaluation of CAD schemes that were published in Medical Physics [15][16][17][18][19][20][21] and IEEE Transactions on Medical Imaging [22][23][24] in 2004. We found that, among the 10 papers we reviewed, nine employed incorrect evaluation methods leading to increased bias and/or variance in the estimated performance levels [9].…”

3D Ultrasound Image Segmentation: Interactive Texture-Based Approaches

“…The authors conclude that the use of complementary and/or redundant texture features allows a more robust segmentation. Sahiner et al (2004) characterize breast masses on three-dimensional ultrasound images. They developed 2D and 3D active contour models for an automated segmentation.…”