B-Flow Sonographic Evaluation of Hemodialysis Fistula

B-Flow Sonographic Evaluation of Hemodialysis Fistula


Cem Yücel, MD ; Suna Özhan Oktar, MD ; Yasemin Erten, MD ; Aykut Bursali, MD ; Hakan Özdemir, MD

1 Departments of Radiology (C.Y., A.Ö.O., A.B., H.Ö.) and Internal Medicine (Y.E.), Gazi University School of Medicine, Ankara, Turkey.

OBJECTIVE: The purpose of this study was to determine the effectiveness of B-flow sonography in the evaluation of hemodialysis fistulas and to compare this new technique with color and power Doppler sonography.

METHODS: In this study, 32 randomly selected patients (mean age, 46 years; age range, 18–87 years) with normally functioning hemodialysis fistulas were evaluated with low– and high–pulse repetition frequency (PRF) color and power Doppler sonography (PRF values of 3 and 10 kHz) and B-flow sonography. All images were reviewed and graded independently by 2 observers for luminal filling with flow signals, visibility of the intimal layer, and overall image quality. The study was approved by the Institutional Review Board, and informed consent was obtained from all patients.

RESULTS: Statistical analysis with Friedman and Wilcoxon signed rank tests revealed that B-flow sonography was superior to other techniques for luminal filling and visibility of the intimal layer (P = .000). For overall image quality, B-flow sonography was also the best method according to the Friedman test (P = .000). However, the Wilcoxon test showed no significant difference between B-flow and high-PRF (10-kHz) color and power Doppler sonography (P = .131). The scores reflected moderate to good interobserver agreement (= 0.285–0.784).

CONCLUSIONS: B-flow sonography is a relatively new and superior imaging technique that provides direct visualization of the blood echoes and the morphologic characteristics of the surrounding vessel wall simultaneously. During the evaluation of hemodialysis shunts, the major advantage of this technique is its ability to avoid artifacts such as aliasing and overwriting.