Device Motion Indication: A Novel Image-Based Tool to Measure Relative Device Motion During Coronary Intervention

Abstract: Background. Movement of the stent delivery system in the coronary bed as a result of the cardiac cycle is a well-known clinical observation that usually is either underestimated or ignored. This effect may potentially jeopardize precise stent deployment. We used a novel technology to objectively measure the relative intracoronary device motion in the different coronary segments throughout the cardiac cycle. Methods. A total of 193 patients undergoing coronary angiography were enrolled and their studies were analyzed for device movement using the SyncVision System (Philips Volcano). The SyncVision System is an add-on image-processing system with unique enhancement and stabilization capabilities. A new feature, the device-motion indication, can precisely display the predeployment device movement measurement (DMM). The device movements within the three segments of each coronary artery were recorded. Results. We identified 218 branch point sites. Median axial displacement was 2.97 mm at the right coronary artery (RCA), 2.22 mm at the left circumflex, and 1.84 mm at the left anterior descending segments. The most movable segments were mid and distal RCA (P<.05). Both heart rate and cardiac contractility significantly affected DMM. Conclusions. The study demonstrates an innovative feature of the SyncVision System as a tool to precisely measure relative device displacement. We claim that the relative device movement is an important quality metric to consider for achieving an effective stent implantation process.  

J INVASIVE CARDIOL 2017;29(12):421-424. Epub 2017 September 15.

Key words: geographical miss, stent deployment, thrombosis, restenosis

The coronary arteries are curvilinear structures that undergo large dynamic variations during each cardiac cycle due to their position on the beating heart, as their displacement tracks that of the underlying epicardium. These orchestrated movements even led Sun and Zhou¹ to suggest the extraction of myocardial dynamic information by estimating coronary arterial motion field. In the cardiovascular arena, percutaneous catheter-based precise stent implantation is a fundamental quality metric with high impact on short-term and long-term sequelae. Yet, extensive relative coronary movements may compromise accurate stent landing. 

The new SyncVision System (Philips Volcano) is an add-on image processing system with unique enhancement and stabilization power. A novel feature of this system, the device motion indication (DMI), can detect the intraluminal device in the x-ray image stream and precisely measure its relative axial and radial movements prior to stent deployment, in an enhanced and stabilized background.

The objective of the current study was to investigate the extent of device movements in the different coronary arteries and their segments, and discuss its practical applications. 

Methods

Study design. The initial estimated enrollment was 250 patients. However, on the predetermined final data collection date, 193 patients who underwent percutaneous coronary intervention (PCI) for accepted clinical indications were recruited (Table 1). At this stage, we performed interim statistical analysis and found meaningful differences considering the primary outcome measures; thus, we terminated the study at this stage. The study protocol was approved by the local Institutional Review Board at the Hillel Yaffe Medical Center in Hadera, Israel. All angiographic procedures were simultaneously processed using the SyncVision System. Patients were excluded if they had unprotected left main disease, target stenosis located in a saphenous vein or in an arterial bypass graft, or in the case of a short life expectancy (≤12 months). Patients were followed for 12 months post intervention. Patient contact was through an office visit or telephone interview. The study was registered (NCT01816347). 

Description of technology. The study made specific use of the regulatory cleared (Food and Drug Administration, CE Mark) SyncVision System and its DMI feature. The system was preinstalled in the site’s cath lab and was used on a day-by-day basis.

The SyncVision is an image acquisition and processing workstation. It captures the fluoroscopic image stream on-line and provides the operator with: (1) angiogram selection, quantitative coronary analysis (QCA), and vessel-region enhancement performed instantly and online during lesion evaluation; and (2) an online enhanced image stream displayed side-by-side with the existing fluoroscopic image stream during device positioning and deployment, as well as post deployment.

All functions performed by the SyncVision are presented, both in the procedure cath lab and in the control room, on a display situated side-by-side with the already existing display of the fluoroscopic image stream. 

DMI feature description. The DMI image is an enhanced view of the vessel as seen during contrast injection, where yellow dots represent the locations where the device’s radio-opaque markers have been detected, using the vessel’s anatomy as an anchor. Dark blue dots represent the marker locations at end-diastole phase (Figure 1).

The objective of such a feature is to enable a better view of the locations in which the device markers have been tracked during positioning injections, and thus facilitate the evaluation of the device location with respect to the lesion to be treated. In cases of a low-quality image, too short injection (<5 sec), or unclear or small markers, the DMI image generation may not be possible. 

Detailed description. The motion of a device relative to the vessel segment throughout the motion cycle(s) of the vessel is depicted over a single enhanced image. The system generates this enhanced image by aligning consecutive image frames onto a selected end-diastole “template” image and then averaging these images among the aligned frames.

Angiographic analysis. Angiography at 20 frames/sec was utilized for image acquisition. The device movement measurement (DMM) was obtained after manual injection of 3-5 mL non-ionic contrast agent under a flow rate of 3 mL/sec. Each scene was completed without any table movement. The coronary segments were divided adopting the modified American Heart Association classification (Figure 2).

The DMM acquisition was achieved simultaneously during routine angiography. There is no intraobserver or interobserver variability in the software.

Effect of various parameters on DMM. We investigated the impact of the specific coronary artery examined, lesion length (<20 mm vs >20 mm), left ventricular contractility (assessed by echocardiography), blood pressure (systolic <140 mm Hg vs >140 mm Hg; diastolic <80 mm Hg vs >80 mm Hg), gender, and heart rate (<80 beats/min vs >80 beats/min) on the extent of axial and radial displacement as assessed by the SyncVision System. 

Statistical analysis. Statistical analysis was performed by IBM Statistics v. 20. Quantitative data are presented as mean ± standard deviation/error or as median values. Data comparisons for statistical significance were assessed using Student’s unpaired t-test. The Bonferroni test was applied for multivariate analysis. Significant differences were defined when the P-value was <.05 using a 2-tailed test.

Results

Using the SyncVision System, we were able for the first time to conclusively confirm substantial device displacement during real-time cineangiography. We found considerable device oscillation during the cardiac cycle with a large variance between the different branch sites. 

Axial device displacement. In 193 patients, a total of 218 sites (1.13 ± 0.4 sites/patient) were identified and used to quantify balloon displacement. Ninety-eight left anterior descending (LAD) segments, 49 left circumflex (LCX) segments, and 71 right coronary artery segments were evaluated. The imaging site was proximal in (65%), mid in (30%), and distal in (5%). Median axial displacement as presented in Table 2 was 2.98 ± 0.45 mm for the RCA, 2.22 ± 0.4 mm for the LCX, and 1.84 ± 0.23 mm for the LAD segments. Axial displacement values were higher in the mid (3.47 mm) and distal (4.49 mm) RCA relative to its proximal part (1.90 mm), higher in the proximal LAD artery (2.01 mm) relative to the mid (1.36 mm) and distal parts (1.79 mm), and higher in the mid and distal LCX artery (2.56 mm) relative to its proximal part (1.71 mm). The most displaced segment was the distal RCA, followed by its mid segment.

Radial device displacement. The median radial displacement was 0.8 mm for the RCA, 0.56 mm for the LCX, and 0.53 mm for the LAD artery (Table 3). Balloon oscillation in the radial plain showed the same trend as in the axial one, being highest in the RCA followed by the LCX and LAD coronary arteries. 

Effect of blood pressure and heart rate. Mean displacement values for systolic blood pressure >140 mm Hg and <140 mm Hg were 1.91 mm and 2.38 mm, respectively (P=.02). Mean values for axial movement for heart rate >80 beats/min and <80 beats/min were 2.02 mm and 2.41 mm, respectively (P=.04). There was no statistically significant effect of diastolic blood pressure on axial movement. Although the radial motions showed a similar trend, they were not statistically powered (0.65 mm vs 0.60 mm; P=.49).

Effect of lesion length. This parameter played no effect on axial displacement. However, at the radial plain, the longer the lesion, the higher the displacement obtained (0.65 mm; P<.05).

Effect of left ventricular function. The mean axial displacement was 2.45 mm with normal and mild left ventricular (LV) dysfunction compared to 2.11 mm at moderate and severe LV dysfunction (P<.01). The radial motions showed the same pattern, ie, 0.62 mm vs 0.46 mm (P<.01).

The effect of gender and coronary dominance. Neither parameter was found to be associated with balloon displacement in both plains.

Multiregression analysis for all parameters. Applying the Bonferroni test, we found the distal RCA to be the most axially displaced segment, followed by the mid RCA. Our results clearly showed that both heart rate and LV contractility significantly affected the relative displacement; therefore, device displacement was higher with lower heart rate and better cardiac function. 

Discussion

This is the first study to conclusively demonstrate the presence of axial displacement of the balloon throughout the cardiac cycle. The displacement was unique for each of the three epicardial coronary arteries, with the RCA and LCX exhibiting larger cyclic displacements than the LAD artery, most likely related to their distinct anatomical courses located in the atrioventricular groove and subjected to substantial lateral displacement induced by atrial contraction. The RCA displacement was approximately 1.34 times that of the LCX, and 1.61 times that of the LAD. Although we present the change of device position in relation to the artery as device displacement, we realize that it is more likely that the vessel moved around the device during the cardiac cycle, with the device probably fixed in position.

Direct measurements of coronary motion were described previously using different imaging modalities^2-4 including biplane or single-plane x-ray and magnetic resonance imaging (MRI), which demonstrated a highly displaced RCA and LCX relative to the LAD artery. Our results are in line with these studies. 

In virtual coronary angiography, motion variability throughout the cardiac cycle can potentially be mitigated, and high-resolution images guaranteed by choosing a small acquisition window (<120 msec).⁵ The optimal MRI period for acquisition was observed at mid diastole after the rapid filling phase, which is actually the period with minimal coronary motion. 

The present study clearly documented that both heart rate and LV contraction significantly affect device oscillation. A slow heart rate gives the illusion of a steadier platform; however, diastole prolongation is accompanied by augmentation of the fractional shortening, ending with a significantly higher displacement. Atsunori et al⁶ supported this observation by showing that rapid ventricular pacing is a safe and efficient means to overcome extensive displacement of the stent delivery system and enhance precise stent deployment. The same analogy is relevant to LV contraction. None of the other factors evaluated significantly affected device displacement. 

Although the SyncVision System is provided with the IVUS and angiography co-registration feature, which can potentially enhance landing-zone assessment, this tool does not take into account the existence of relative device displacements. We believe that device displacement is a modifiable factor that may potentially play a significant role in the occurrence of geographical miss. 

Clinical implications. Rigorous and precise stent deployment is of paramount importance, critically impacting both short-term and long-term sequelae. Sakurai et al⁷ demonstrated the importance of IVUS-guided full lesion coverage to avoid edge stenosis. Incomplete stent landing occurs in 20%-30% of cases and it is realized much more if stent expansion is assessed by IVUS.⁸ 

In real life, the lesion dimension estimations are mostly subjective and inaccurate, and the working platform is dynamic rather than static. These two issues are adequately addressed by the QCA and DMI features in the SyncVision System.

Current PCI methodology addresses neither the timing of stent deployment during the cardiac cycle nor the occurrence of device displacement throughout the different coronary segments. In virtual coronary angiography, electrocardiogram (ECG)-gated acquisition for three-dimensional image reconstruction⁹ may help to circumvent this drawback driven by the continuously displaced arteries; the potential concept of ECG-gated balloon inflation might be a promising means to overcome this cycle-dependent device oscillation.

Conclusion

Using the DMI feature of the SyncVision System, we were able for the first time to objectively obtain and measure device displacement at the different coronary artery segments. We assume that this factor is relevant in the occurrence of geographical miss, and should thoroughly be accounted for when determining stent length. The potentially negative impact of device displacement on geographical miss can basically be mitigated by adding extra stent length equivalent to the DMM values. 

References

1.    Sun Z, Zhou Y. Assessing cardiac dynamics based on x-ray coronary angiograms. J Multimedia. 2013;8:48-55.

2.    Achenbach S, Ropers D, Holle J, et al. In-plane coronary arterial motion velocity: measurement with electron-beam CT. Radiology. 2000;216:457-463.

3.    Shechter G, Resar JR, McVeigh ER. Displacement and velocity of the coronary arteries: cardiac and respiratory motion. IEEE Trans Med Imaging. 2006;25:369-375.

4.    Hofman MB, Wickline SA, Lorenz CH. Quantification of in-plane motion of the coronary arteries during the cardiac cycle: implications for acquisition window duration for MR flow quantification. J Magn Reson Imaging. 1998;8:568-576.

5.    Dewan M, Hager GD, Lorenz CH. Image-based coronary tracking and beat-to-beat motion compensation: feasibility for improving coronary MR angiography.  Magn Reson Med. 2008;60:604–615.

6.    Okamura A, Ito H, Iwakura K, et al. Rapid ventricular pacing can reduce heart motion and facilitate stent deployment to the optimal position during coronary artery stenting: initial experience. EuroIntervention. 2007;3:239-242.

7.    Sakurai R, Ako J, Morino Y, et al; SIRIUS trial investigators. Predictors of edge stenosis following sirolimus-eluting stent deployment (a quantitative intravascular ultrasound analysis from the SIRIUS trial). Am J Cardiol. 2005;96:1251-1253.

8.    Calvert PA, Brown AJ, Hoole SP, et al. Geographical miss is associated with vulnerable plaque and increased major adverse cardiovascular events in patients with myocardial infarction. Catheter Cardiovasc Interv. 2016;88:340-347. Epub 2015 Nov 3.

9.    Achenbach S, Ulzheimer S, Baum U, et al. Noninvasive coronary angiography by retrospectively ECG-gated multislice spiral CT. Circulation. 2000;102:2823-2828.

From the Department of Cardiology, Hillel Yaffe Medical Center Affiliated with Rappaport Faculty of Medicine, Technion – Israel Institute of Technology, Haifa, Israel. 

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Prof. Frimerman reports that he is a consultant to Sync Rx, Netanya, Israel. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted March 28, 2017, provisional acceptance given April 6, 2017, final version accepted May 22, 2017.

Address for correspondence: Rami Abu Fanne, MD, PhD, Hashalom Street, Hadera, 38100, POB 169 Israel. Email: ramia@hadassah.org.il