藤原 良将¹, 久 直史², 吉田 祥二¹

Yoshimasa FUJIWARA¹, Naofumi HISA², Syoji YOSHIDA¹

¹高知医科大学放射線科, ²高知医科大学放射線部

¹Department of Radiology Kochi Medical School, ²Department of Radiology Kochi Medical School

キーワード : Doppler ultrasonography, Flow phantom model, Pulsatility index, Shunt, Stenosis, V<SUB>max</SUB>

Although the organs and abnormal lesions are known to produce different Doppler signals, the significance of these signals remains to be fully analyzed. Here we use a flow phantom model to analyze color Doppler signals with fast Fourier transform (FFT) analysis in hepatic lesions and various other situations. A Cobe-Stöckert circulating pump was used as a pulse generator and silicone tubes were used as vessels. The shunt and stenosis models were joined by control lines. Tubes 4 mm in diameter were used to simulate arterial pulsatile flow in each model. These arterial tubes bifurcated to the control and lesional (stenosis or shunt) lines, which were 2 mm in diameter and which were fixed in a water bath. A Duplex Doppler device was used to measure both pulsatility index and maximum velocity in the water bath. Beyond the points at which these measurements were taken, the tubes were gathered and run into a drainage bag, proximal to which peripheral pressure was monitored. This system was then filled with laboratory Doppler test fluid (ATS, stock number 707-G), which had a viscosity of 1.66±0.1 centistokes. The Reynolds number of the measuring points ranged from 120 to 2,000. Resulting changes in waveform and wave velocities were monitored with a 128×P/10 Duplex Doppler device (Acuson) when the original pulse waveform and pulse rate were varied. Angle of the Doppler beam was fixed at about 60 degrees. Flow data were obtained from a range of situations. Pulsatility index (PI) and maximum velocity (V_max) of each waveform were evaluated. In the stenosis model, Doppler signals were obtained at the prestenotic and stenotic sites. V_max and PI were measured under various degrees of peripheral pressure and peripheral stenosis at the prestenotic and stenotic sites. Stenosis ratio was about 50%. In the shunt model, measurement was made at preshunt and shunt sites. The shunt size was 0.5 mm in diameter. Color-flow imaging and FFT analysis of each model were performed 5 to 10 times for each measuring point, and the Mann-Whitney test was used for statistical analysis. Mean V_max at stenotic sites was 67.0±21.6 cm/sec (mean±SD), significantly higher than that of the control line (control=39.5±2.30 cm/sec) (95% significance). Mean prestenotic V_max was 32.6±2.07 cm/sec, lower than that of the control line. V_max of the shunt and preshunt sites was 238±47.0 cm/sec and 56.2±6.30 cm/sec, respectively, significantly higher than that of the base line (95% significance). Mean PI at stenotic sites was 7.42±0.71 (mean±SD), within the same range as that of the control line (control=7.55±0.67). PI values at shunt and preshunt sites were 2.82±1.58 and 5.46±0.61, respectively, significantly lower than that of the control line (control=7.55±0.67). V_max and PI at shunt sites changed continuously with increase in peripheral pressure. Increases in V_max were proportional to increases in shunt pressure gradient, and decrease in PI was inversely proportional to increase in peripheral pressure. Regression curves for the respective groups were y=l.65+2.0x and y=5.37-0.07x. Peak of the systolic stroke at the stenotic site decreased slightly as peripheral pressure increased in the stenosis models, and the reverse-flow component at diastole was apparent, as was increase in PI. Decrease in V_max and PI was proportional to increase in rate of peripheral stenosis when the control line was clamped. PI increased significantly in severe stenosis, however. We thus conclude that both stenosis and shunt produce high-velocity signals; decrease in PI occurs at preshunt and shunt sites; and PI does not change significantly at stenotic sites. Although high PI may have been discernible in severely stenotic lesions, high-velocity flow with PI is probably related to shunt flow. These observations should lead to a better understanding of Doppler signals with FFT analysis in vivo.