In vitro validation of new Fourier-domain optical coherence tomography

Abstract

To validate the accuracy and evaluate the intra- and inter-observer variability of vascular measurements using novel Fourier-domain optical coherence tomography (FD-OCT) versus intravascular ultrasound (IVUS) in a coronary phantom model. A dedicated cylindrical phantom comprised of four sections with differing dimensions was used to compare images generated by five FD-OCT and five IVUS catheters. Each FD-OCT catheter was tested using three different consoles, generating 15 imaging pullbacks. Two independent experts, blinded to the phantom data, performed measurements. There were 180 FD-OCT and 60 IVUS cross-sectional measurements of mean lumen diameter (MeanLD), minimal lumen diameter (MLD), maximum lumen diameter (MaxLD) and lumen area (LA) at three points in each section. FD-OCT measurements had excellent correlations with IVUS (concordance correlation coefficient [CCC]: ≥0.9769 for MeanLD, LA or longitudinal length [LL]) and phantom dimensions (CCCs for FD-OCT: ≥0.9958 for MeanLD, LA or LL). FD-OCT measurements were larger than IVUS (p<0.0001), but showed less measurement errors compared to IVUS (p<0.0001). Moreover, FD-OCT caused less discrepancy between MaxLD and MLD versus IVUS (p<0.0001). Intra- and inter-observer variability was low for both FD-OCT (CCCs for MeanLD, LA and LL ≥0.9996) and IVUS (≥0.9935). Image catheter position did not influence FD-OCT measurements. FD-OCT was more accurate than IVUS and had similar high reproducibility to determine vascular dimensions in vitro. These results support the use of FD-OCT in the clinical setting.