A sensitive new indicator for diagnostics of ovarian malignancy, based on the Doppler velocity spectrum

Ovarian cancer is the most deadly gynecological malignancy, with an overall survival rate of about 35%. Approximately 60% of the cases of ovarian cancer are lethal. Ultrasonic examination, including Doppler imaging, is a commonly used technique for the diagnosis of ovarian masses. Two major clinical parameters, currently derived from the Doppler flow waveform, are the resistance index (RI) and the pulsatility index (PI). The decay constant of the Doppler waveform, which characterizes its decrease from systole to diastole as an exponential decay, has recently been presented as an additional measure of tumor malignancy. In this paper, we have analyzed the velocity spectrum of the Doppler flow signal to determine if it reveals differences that might contribute to the diagnosis of malignancy. We designed a new parameter characterizing the slope of the mean velocity spectrum at end-diastole ("End Diastolic Velocity Distribution Slope," referred to as DVD_S). Additional indices, related to various approaches for the analysis of the Doppler image, were also derived. However, they proved to be inferior to the DVD_S. The DVD_S was tested on 20 benign and 33 malignant ovarian images. This new parameter seems to provide a good ability to discriminate between the two types of tumor. Its mean value is 1.90 +/- 1.33 for malignant tumors, compared to 9.21 +/- 5.34 for benign masses (area under ROC curve: 0.983), yielding a detection rate of about 94%. In fact, this parameter provides much better results than the previously used variables, and has the potential to significantly improve the detection of malignancy.