Effect of quantitated stenoses on arterial sound spectrum analysis

Abstract

Spectral analysis of carotid Doppler signals has been shown to be accurate in determining the degree of carotid artery stenosis when compared with arteriography. Since the Doppler angle and the amount of gain used during measurement may affect results, an in vitro study was performed measuring peak frequency, mode frequency, and percent window at various Doppler angles and gains to determine if these latter factors could affect results of the test. A spectrum analyzer with an 8 MHz continuous-wave Doppler probe was utilized. Seven cross-sectional area stenoses (0 to 100 percent) were applied to an undiseased bifurcated cadaver artery which was suspended in a saline bath and placed in a pulsatile circulatory system. At each stenosis, peak frequency, mode frequency, and percent window were measured at three Doppler angles (45 degrees, 60 degrees, and 75 degrees) at a constant gain and at three different gains (low, medium, and high) at a constant Doppler angle of 60 degrees. Arterial pressure was measured distal to the stenoses. Critical stenoses were present at greater than 82 percent area reduction. At a Doppler angle of 60 degrees and medium gain, correlation coefficients between percentage of stenosis and peak frequency, mode frequency, and percent window were 0.9520, 0.8369, and -0.9861, respectively. However, peak frequency and mode frequency actually decreased at stenotic areas of more than 82 percent and were similar to frequencies seen at stenotic areas of 61 percent. Peak frequency significantly decreased as Doppler angle increased, so there was an even greater overlap of peak frequency values at larger angles. Percent window did not appear to be affected by Doppler angle. At an angle of 60 degrees and different gains, there appeared to be very little overlap of percent window values between any stenoses. Peak frequency and mode frequency did not appear to be affected by gain. This study demonstrates an excellent correlation between percentage of stenosis and peak frequency and percent window. However, peak frequency was significantly affected by changes in Doppler angle and did not differentiate subcritical (61 percent) from critical (96 percent) stenoses. Percent window, however, was not significantly affected by Doppler angle or gain and was able to differentiate between all degrees of stenosis.