Evaluation of the influence of cardiac motion on the accuracy and reproducibility of longitudinal measurements and the corresponding image quality in optical frequency domain imaging: an ex vivo investigation of the optimal pullback speed.

Longitudinal measurement using intravascular ultrasound is limited because the motorized pullback device assumes no cardiac motion. A newly developed intracoronary imaging modality, optical frequency domain imaging (OFDI), has higher resolution and an increased auto-pullback speed with presumably lesser susceptibility to cardiac motion artifacts during pullback for longitudinal measurement; however, it has not been fully investigated. We aimed to clarify the influence of cardiac motion on the accuracy and reproducibility of longitudinal measurements obtained using OFDI and to determine the optimal pullback speed. This ex vivo study included 31 stents deployed in the mid left anterior descending artery under phantom heartbeat and coronary flow simulation. Longitudinal stent lengths were measured twice using OFDI at three pullback speeds. Differences in stent lengths between OFDI and microscopy and between two repetitive pullbacks were assessed to determine accuracy and reproducibility. Furthermore, three-dimensional (3D) reconstruction was used for evaluating image quality. With regard to differences in stent length between OFDI and microscopy, the intraclass correlation coefficient values were 0.985, 0.994, and 0.995 at 10, 20, and 40 mm/s, respectively. With regard to reproducibility, the values were 0.995, 0.996, and 0.996 at 10, 20, and 40 mm/s, respectively. 3D reconstruction showed a superior image quality at 10 and 20 mm/s compared with that at 40 mm/s. OFDI demonstrated high accuracy and reproducibility for longitudinal stent measurements. Moreover, its accuracy and reproducibility were remarkable at a higher pullback speed. A 20-mm/s pullback speed may be optimal for clinical and research purposes.