Internal validation of an automated system for brachial and femoral flow mediated dilation

Abstract

Background: Flow Mediated Dilation (FMD) has immense potential to become a clinical, non-invasive biomarker of endothelial function and nitric oxide bioavailability, which regulate vasomotor activity. Unfortunately, FMD analysis techniques could deviate significantly in different laboratories if a validation process is not involved. The purpose of this study was to provide validation to the assessment of FMD analysis in our laboratory and to standardize this process before reporting results of FMD.

Methods: Brachial and femoral arteries FMD were performed on 28 apparently healthy participants (15 male and 13 female, ages 18-35 years). For the intratester reliability study, nine subjects were asked to come to the lab for a second brachial FMD within 48 h. All FMD procedures were performed by the same investigator, while the FMD analyses were performed by 2 independent testers who were blind to each other's analyses. FMD analyses included baseline artery diameter measurements, peak artery diameter after 5 min of ischemia, and FMD. Analysis was completed via an automated edge detection system by both testers after training of the methodical process of analysis to minimize variability. Intratester and intertester reliability were determined by using coefficient of variation (CV) between first and second visit (intratester) and between results obtained by both testers (intertester).

Results: The intratester CVs for tester 1 and 2 were 3.28 and 2.62%, 3.74 and 3.27%, and 4.95 and 2.38% for brachial baseline artery diameter, brachial peak artery dilation, and brachial FMD, respectively. In the intertester CVs were 2.40, 3.16, and 3.37% for brachial baseline artery diameter, peak artery dilation, and FMD, respectively and 4.52, 5.50, and 3.46% for femoral baseline artery diameter, peak artery dilation, and FMD, respectively.

Conclusion: All CVs were under or around 5%, confirming a strong reliability of the method. Our laboratory has shown that the FMD protocol is reproducible due to the significantly low coefficient of variation. This is one step closer to use FMD as a biomarker for endothelial function in our laboratory.

Keywords: Coefficient of variation; Endothelial function; Flow mediated dilation; Validation.

Fig. 1

Screenshot of a representative brachial FMD analysis using edge-detection technology. On the left-hand panel is the analyzed frame from the ultrasound imaging system and on the right-hand panel are the actual measurements from frame 9 to 27, including a time-based graph on the overall FMD on the right-top-panel. On the middle, the region of interest (ROI) has been zoomed for better view