Frequency analysis of shunt sounds in the arteriovenous fistula on hemodialysis patients

摘要

The aim of this research is to evaluate vascular access (VA) using shunt sounds. We measured the blood flow volume in the brachial artery (FV) and the resistance index (RI) by using ultrasonic diagnostic equipment, and the frequency and spectrum characteristic to the shunt sounds by using an electronic stethoscope. The subjects were fifty hemodialysis patients who had been treated with arteriovenous fistura (AVF). An electronic stethoscope was used for all the subjects to record the shunt sounds of arteriovenous anastomosis with an 8 kHz sampling frequency for 8 seconds. The digitized sound was analyzed by the Fourier transform, and every 10 Hz spectra were examined for acoustic amplitude and spectral energy distribution. Based on the ultrasonographic data, subjects FV >= 500 mL/min) and II (FV < 500 mL/min), and Groups A (RI < 0.6) and B (RI >= 0.6). Spectra distribution on frequency bands were compared between two different groups. The spectra in 41–270 Hz were significantly higher in Group I than in Group II. In Group I, the spectra in 111–160Hz were significantly higher than those in other frequency bands. In Group II, the spectra in 121–140 Hz were significantly higher than in the other frequency bands. Comparing Groups A and B, Group A showed a significantly higher spectrum in the 51–270 Hz, and in particular, the spectra of 121–160 Hz showed significantly higher compared with those of the other frequency bands. In group B, spectra of 131–140 Hz showed significantly higher compared with those of the other frequency bands. From these results, shunt sounds acquired in patients with good AVF function (FV >= 500 mL/min or RI < 0.6) were larger in the spectra of a 51–270 Hz frequency bands and were broad in the frequency domain compared with those in patients with poor AVF function (FV < 500 mL/min or RI >= 0.6). I- was found that analyzing the amplitude and frequency features of spectra in shunt sounds would be used for identifying FV and RI.