Feasibility of alternative trigger deletion to improve 64-slice coronary computed tomography angiography quality in patients with increased heart rate

Introduction

Coronary computed tomography angiography (CCTA) has received a Class I, evidence Level A recommendation as a substitute for conventional invasive coronary angiography (ICA) in patients with suspected coronary artery disease (CAD) due to its high accuracy in detecting significant coronary stenosis (1-4). In addition to CCTA-based plaque evaluation, artificial intelligence-derived CCTA parameters, such as the fractional flow reserve and fat attenuation index, are crucial for enhancing risk stratification and guiding therapeutic strategies for patients with CAD (1-7).

CCTA success relies on multiple factors, among which, the heart rate (HR) is crucial. Rapid cardiac motion and heart rate variability (HRV) can cause motion artifacts, affecting image quality and diagnostic accuracy (8). To address this issue, hardware-based approaches (e.g., faster gantry rotation speed, wider detector coverage, and the dual-source and photon-counting technique) (9-12) and software-centric solutions (e.g., motion correction algorithm and artificial intelligence) have been developed (13-16). The limited temporal resolution of conventional multi-detector computed tomography (MDCT) scanners can lead to substantial motion artifacts, affecting CCTA performance (17,18). Achieving the target HR before CCTA remains essential for optimal 64-slice CCTA image quality, requiring HR-lowering medications (19). Further, maintaining stable HR during the scan is challenging due to breath-holding and contrast-induced warmth (19,20). An elevated HR will lead to accelerated vascular motion velocity, while simultaneously causing a mismatch between the preset scan pitch and the actual HR. These two factors ultimately contribute to degraded image quality (19,21).

Retrospective electrocardiogram (ECG)-gated helical acquisition enables full cardiac cycle volumetric data capture, allowing for the arbitrary adjustment of the position of the temporal windows for image reconstruction while also permitting the ECG-edit function to correct artifacts caused by high HR or HRV (22,23). Selecting an optimal reconstruction phase using the relative phase (%R-R) technique becomes challenging when the HR is highly variable. Reconstruction based on absolute timing (R+ absolute time, R− absolute time) is carried out at a fixed interval following the previous R peak or preceding the next R peak, making it less susceptible to fluctuations in HR (23). The ECG-edit procedure involves arbitrary modifications of the R-wave trigger to supply the reconstruction software with information that minimizes the residual motion. Previous studies (22,24-29) used the ECG-edit function to delete short R-R intervals that hindered the selection of consistent cardiac phases from the ECG data, resulting in clinically acceptable images in patients with arrhythmias.

Previous studies on the ECG-edit function were limited to patients with abnormal heart rhythms. Nevertheless, the method of deleting abnormal triggers inspired us to hypothesize that in patients with regular heart rhythms, alternating ECG triggers could mimic the effects of abnormal triggers like those in premature atrial contractions or premature ventricular contractions. By deleting these “abnormal triggers”, we speculate that image quality could be improved in patients with increased HR. Therefore, our study aimed to evaluate the feasibility and effect of deleting ECG triggers alternately on improving the quality of 64-slice CCTA images, and identify the optimal strategy for using this novel ECG-edit function for clinically acceptable images. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-897/rc).

Methods

Patients

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Bishan Hospital of Chongqing Medical University (No. cqbykyll-20250110-2, dated January 10, 2025), and individual consent for this retrospective analysis was waived. The sample size was selected based on the approach described by Kaniewska et al. (30). We consecutively enrolled 102 patients who underwent CCTA for known or suspected CAD between January 2021 and April 2024. We included patients with a HR <70 beats per minute (bpm) before the scan but ≥75 bpm during the scan (including cases with sustained HR elevation or a stable high HR). The exclusion criteria included prior cardiac implantable electronic device placement, coronary stent implantation, or coronary artery bypass grafting, and contraindications to CCTA, including a history of hypersensitivity to the iodinated contrast agent, impaired renal function (creatinine clearance <60 mL/min), an inability to maintain a 10-s breath-hold, and pregnancy. Overall, 49 patients were excluded from the study: 26 due to previous stenting, 14 due to a previous history of coronary bypass grafting surgery, and 9 due to the presence of a pacemaker. Thus, ultimately, 53 patients were included in this study. Figure 1 provides the patient enrollment flowchart.

Figure 1 Flow chart showing the inclusion and exclusion criteria and patient selection workflow. CAD, coronary artery disease; CCTA, coronary computed tomography angiography; HR, heart rate; bpm, beat per minute.

Image acquisition

The computed tomography (CT) scans were obtained with a 64-slice CT scanner (Discovery CT 750 HD, GE Healthcare, Milwaukee, WI, USA) using the following parameters: detector collimation: 64×0.625 mm; gantry rotation time: 0.35 s, and pitch: 0.20. The tube voltage was set at 120 kVp with the Smart-mA ranging from 220 to 600 mA according to the patient’s body mass index (BMI). ECG dose modulation, which reduces the current for nontarget phases, was used in the data acquisition. The scan range extended from 10 mm below the carina to 10 mm below the dome of the diaphragm. All the acquisitions were retrospectively gated with craniocaudal acquisition during sustained inspiratory apnea. All the images were reconstructed with a display field of view of 200 mm, and a slice thickness and interval of 0.625 mm. Data acquisition was triggered by a bolus-tracking technique, which started 8.0 s after the ascending aorta attenuation surpassed the predefined threshold of 100 Hounsfield units (HU). Contrast agent (50–60 mL; 350 mg iodine/mL, Iohexol, Yangtze River Pharmaceutical Group, China) followed by saline solution (30–35 mL) was continuously injected through the right antecubital vein at a flow rate of 4.5–5.0 mL/s using an 18-gauge catheter. The dose and flow rate of the contrast agent were determined based on each patient’s BMI and vein condition.

Image reconstruction and ECG-edit function

The axial images were reconstructed using 50% Adaptive Statistical Iterative Reconstruction (ASiR-Slice, GE Healthcare) with the standard reconstruction type. Two datasets were established based on whether the ECG-edit was performed, with the reconstruction strategies tailored to each patient’s HR to minimize motion artifacts. The first dataset maintained original ECG signals without applying the ECG-edit function, while the second dataset underwent collaborative refinement by dual-certified cardiac imaging specialists employing the ECG-edit function.

Reconstruction was performed in the 2% steps of the R-R interval to select the cardiac phase with minimal motion. If motion artifacts were observed in vessel segments on the initially reconstructed images, a preferable phase to reduce vessel fuzziness was pursued as the optimal phase. The optimal reconstruction phase demonstrating minimal motion artifacts was jointly selected by a cardiovascular radiologist and a radiology technologist through consensus assessment. Among multiple cardiac phase specification methods for image reconstruction, including R+ absolute time, R− absolute time, and relative phase (%R-R), individualized selection was performed based on case-specific anatomical and physiological considerations.

In the first dataset, in patients with regular rhythm and a HR exceeding 75 bpm, progressive shortening of the mid-diastolic phase shifted the optimal reconstruction window to end-systole, requiring relative phase (%R-R) approaches (19,31). For arrhythmic patients, where conventional ECG-derived phase selection became unreliable, the absolute timing methods demonstrated superior utility for identifying motion-minimized cardiac phases compared to the relative phase approaches (32).

In the second dataset, two distinct ECG-edit strategies were applied based on the alternate deletion starting from either the first or the second trigger point. After reconstructing the images using a step size of 2% of the algorithm-input R-R interval, the optimal images with the least motion artifacts from each strategy were selected for further comparison and analysis. The ECG-edit strategy and reconstruction phase resulting in the fewest artifacts were ultimately determined.

Image analysis

All the images were anonymized and reviewed on a single vendor workstation (AW 4.6, GE Healthcare), and the images were transferred to ShuKun DIGITALBODY Technology Platform (Shukun Technology, Beijing, China) to evaluate the subjective image quality for each patient.

A randomized blinded evaluation protocol was employed by dual-board-certified cardiovascular radiologists (with 4 and 9 years of experience in cardiac CT, respectively) to objectively assess all the reconstructed images. Using a systematic coronary segmentation analysis, each patient’s vascular anatomy was methodically graded across the following 10 standardized regions: left main (LM) coronary artery; proximal/mid/distal left anterior descending (LAD) artery; proximal/distal left circumflex (LCX) artery; major obtuse marginal (OM) branch or ramus intermedius; and proximal/mid/distal right coronary artery (RCA). The evaluators were strictly blinded to the clinical metadata and technical parameters (33).

A four-point Likert scale system was used to evaluate the subjective image quality on the axial source slices, multiplanar reformations, and curved planar reformations reconstructed from the optimal phase selected by the operators with regard to the degree of motion artifacts on a per-segment level (34). Each segment was ranked as excellent (no artifacts, scored 4), good (minor artifacts, fully evaluable, scored 3), adequate (moderate artifacts, acceptable for diagnosis, scored 2), or poor (severe artifacts impairing accurate evaluation, scored 1) (Figure 2). Any disagreements between the two radiologists regarding the evaluation were resolved by consensus after discussion. Coronary segments scoring 2–4 points received diagnostic clearance, while grade 1 segments were excluded from diagnostic consideration. The analytical framework incorporated tri-level hierarchical assessment (patient/artery/segment) through cascade classification logic; any non-interpretable segment triggered the categorical exclusion of its parent vessel and the corresponding patient from the subsequent analyses, with interpretability rates computed as follows: interpretability rate = [(cleared segments)/(total evaluable segments) ×100%] (15).

Figure 2 Representative curved planar reformation images providing examples of the four-point Likert scale based on the degree of motion artifacts of the right coronary arteries. (A) A score of 1 indicates non-interpretable image quality (severe artifacts with inadequate delineation between the lumen and the surrounding tissue); (B) a score of 2 indicates adequate image quality (a moderately blurred vessel, but acceptable for diagnosis); (C) a score of 3 indicates good image quality (minor artifacts and blurring of the vessel margin, fully evaluable); (D) a score of 4 indicates excellent image quality (with the absence of artifacts).

The objective image quality was assessed using a number of parameters. Image noise and contrast density were defined as the standard deviation (SD) and mean, respectively, of the lumen CT attenuation value in a region of interest (ROI) placed in the mid-RCA (35). The signal-to-noise ratio (SNR) was quantified as the coronary arterial luminal attenuation divided by image noise. As the optimal phases of images reconstructed with and without the ECG-edit could occur at different stages of the cardiac cycle, the ROIs placed along the mid-RCA could be located at different positions within the same slice (Figure 3).

Figure 3 A 26-year-old male with a HR range of 70–81 bpm. The reconstruction phases for images reconstructed without the ECG-edit (A,B) and those reconstructed with the ECG-edit (C,D) were 45% and 38%, respectively. The ROIs were placed at the mid-RCA in the same plane but different positions, the measurement parameters of image (B) and image (D) were mean =114.6 HU, SD =38.0 HU; mean =228.7 HU, SD =39.2 HU, respectively. bpm, beats per minute; ECG, electrocardiogram; HU, Hounsfield unit; RCA, right coronary artery; ROI, region of interest.

Radiation dose parameters

The effective radiation dose of CCTA was determined in millisieverts (mSv) using a scanner-specific CT dose index volume. This dose estimation was derived by applying a chest-specific conversion coefficient (K =0.014 mSv·mGy⁻¹·cm⁻¹) to the dose-length product (measured in mGy·cm) (36).

Statistical analysis

The quantitative variables are expressed as the mean ± SD, while the categorical variables are described as the frequency count and percentage. The Wilcoxon signed-rank test was used to compare the image quality scores. The pair-wise McNemar test was applied to evaluate interpretability discrepancies between the two datasets across the per-patient, per-artery, per-segment analyses, and three coronary arteries. The paired t-test was used to test differences in continuous variables with a normal distribution. The Kappa statistic was used to test inter-observer agreement, and the Kappa value was explained as follows: poor for κ ≤0.20, fair for 0.20< κ ≤0.40, moderate for 0.40< κ ≤0.60, good for 0.60< κ ≤0.80, and excellent for κ >0.80 (37). The statistical analyses were performed using SPSS software version 29.0 (IBM, Armonk, NY, USA). A two-sided P value of <0.05 was considered statistically significant.

Results

Baseline characteristics

The 53 included patients had a mean age of 57.57±14.78 years (range, 26.0–97.0 years), and 54.72% were male. The patients had a mean BMI of 22.60±1.94 kg/m², and the average HR during acquisition was 82.26±6.00 bpm with 7.02±7.52 bpm variability. Comprehensive population characteristics and CCTA parameters are set out in Table 1.

ECG-edit function

To assess whether the novel ECG-edit of alternately deleting triggers improved image quality, two methods were used in which the triggers were deleted starting from either the first or the second trigger (Figure 4). In 38 (71.70%) of the patients in the second study dataset, the optimal reconstruction phases were selected from the data with ECG triggers deleted starting from the first trigger.

Figure 4 Graphs showing the two ECG-edit methods deleting ECG triggers starting from the different triggers. ECG report (A) shows a baseline mean HR of 88 bpm (range, 87–90 bpm), measured via R-wave-triggered analysis, with reconstruction performed at the end-systolic phase (52% of the R-R interval). ECG reports (B,C) show the selective deletion of ECG trigger points: ECG report (B) shows the deletion (as indicated by the arrows) starting from the first trigger point, resulting in an algorithm-input HR of 44 bpm based on the remaining R-waves. ECG report (C) shows the deletion (as indicated by the arrows) initiated from the second trigger point, resulting in a mean algorithm-input HR of 44 bpm (range, 44–88 bpm), where the initial trigger point recorded 88 bpm and all subsequent trigger points recorded 44 bpm. bpm, beats per minute; ECG, electrocardiogram; HR, heart rate.

Optimal reconstruction phase

In the first study dataset, in terms of reconstruction, optimal image quality was observed in the end-systolic phase (40–59% R-R interval) in 47 (88.68%) patients, and in the mid-diastolic phase (70–79% R-R interval) or the end-diastolic phase (90–99% R-R interval) in 6 (11.32%) patients. In the second study dataset, in terms of reconstruction, optimal image quality was observed in the systolic phases (20–29% and 70–79% algorithm-input R-R interval) in 46 (86.79%) of the 53 patients, in the diastolic phase (80–99% algorithm-input R-R interval) in 5 (9.43%) patients, and in non-standard cardiac phases in 2 (3.77%) patients. The inter-dataset analysis revealed statistically significant differences in the phase-selection preferences between the two cohorts (P<0.01) (Figure 5), and the 23% reconstruction phase accounted for the largest proportion in the second study dataset. Figure 6 illustrates the relationship among the original HR, the algorithm-input HR, the deleted triggers, and the resulting reconstruction phases. Figure 7 displays images from a representative case.

Figure 5 The distribution of the optimal phases of the two datasets. ECG, electrocardiogram.

Figure 6 Graphs showing the ECG-edit strategy and the optimal reconstruction phase. (A,B) show the original ECG, while (C-F) show the edited ECG. (C,D) were processed by alternately deleting ECG triggers starting from the first trigger, while (E,F) were processed starting from the second trigger. (A,B) show the actual average HR of 76 bpm, while (C-F) show an algorithm-input average HR of 38 bpm (half of the actual HR) calculated based on the remaining ECG trigger after the ECG-edit. (A,B) show that the optimal reconstruction phase corresponds to 45% of the original R-R interval, which aligns with the end-systolic phase of the cardiac cycle. (C,F) show that the optimal reconstruction phase occurs at 75% of the algorithm-input R-R interval, while (D,F) show that it occurs at 23%. As indicated by the yellow dashed lines, the reconstruction windows (as indicated by the blue vertical stripes) in (C,E) correspond to the second, fourth, sixth, and eighth reconstruction windows in (A). Similarly, the reconstruction windows in (D,F) correspond to the first, third, fifth, seventh, and ninth reconstruction windows in (B). bpm, beats per minute; HR, heart rate; ECG, electrocardiogram.

Figure 7 Axial slices (A,D,F) and curved planar reformations (B,E,G) of a 48-year-old male (average HR: 83 bpm). (A,B) The white arrows indicate motion artifacts in the proximal RCA reconstructed without the ECG-edit; (D-G) the white arrows show artifact reduction in the proximal RCA reconstructed with the ECG-edit deleting triggers starting from the first trigger. The original ECG report (C) show that the reconstruction phase was 52%. The ECG reports with triggers deleted starting from the first trigger point (H,I) show reconstruction phases of 23% and 75%, corresponding to the systolic periods of the actual cardiac cycles preceding and following the deleted trigger points, respectively. ECG, electrocardiogram; HR, heart rate; RCA, right coronary artery.

Image quality and interpretability

CCTA subjective image quality

The inter-observer agreement in terms of subjective image quality was good (κ =0.776). The analysis included 530 coronary segments, and the mean subjective quality scores across all the levels of analysis were significantly higher in the second dataset than the first dataset (segment-level: 2.94±0.70 vs. 1.97±0.78; artery-level: 3.18±0.53 vs. 1.93±0.61; patient-level: 3.87±0.34 vs. 2.89±0.64, respectively, all P<0.001). Table 2 sets out the stratified subjective image quality assessments across the segment-, artery-, and patient-level analytical dimensions. Figure 8 displays images from a representative case.

Figure 8 A 83-year-old female with an average HR of 75 bpm during the scan. Axial slice (A) reconstructed without the ECG-edit and the corresponding curved planar reformation image (C) and lumen (D) show motion artifacts (as indicated by the arrows) at the mid-segment of RCA. Axial slice (B) reconstructed with the ECG-edit and the corresponding curved planar reformation image (E) and lumen (F) show no visible motion artifacts (as indicated by the arrows) at the mid-segment of RCA. The ECG report (G) shows that the HR was 79 bpm and the reconstruction phase was 52%. The ECG report (H) shows that the ECG triggers were deleted alternately starting from the first trigger and the algorithm-input HR was 40 bpm with the reconstruction phase of 23%. ECG, electrocardiogram; HR, heart rate; bpm, beats per minute; RCA, right coronary artery.

CCTA image interpretability

The segment-level interpretability analysis revealed that the ECG-edit function significantly improved the interpretability metrics, which was accompanied by a marked decrease in the non-interpretable segments (159/530, 30.00% vs. 20/530, 3.77%, P<0.001). Similarly, the ECG-edited reconstructions demonstrated better performance at both the artery and patient levels, with statistically significant differences in both comparative analyses (142/159, 89.81% vs. 76/159, 47.80%; 37/53, 69.81% vs. 7/53, 13.21%, respectively, both P<0.001). In relation to the three major coronary arteries, significant improvements in interpretability were observed for the RCA and LCX artery (20/53, 37.74% vs. 45/53, 84.91%; 23/53, 43.40% vs. 48/53, 90.57%, respectively, both P<0.001). Detailed comparisons of the image interpretability results between the two datasets are shown in Table 3. Figure 9 displays an example of axial slices reconstructed with and without the ECG-edit.

Figure 9 A 55-year-old man with suspected CAD and a BMI of 23.31 kg/m². Original axial images (A,B) reconstructed without the ECG-edit were non-interpretable due to motion artifacts at the proximal segment (as indicated by the white arrow) of the LAD artery and the mid segment (as indicated by the black arrow) of the RCA. The ECG report (C) shows that the HR was 88 bpm. Images (D-G) and (J-M) were reconstructed with triggers deleted alternately from the first and second trigger, respectively. (D,F,J,L) and (E,G,K,M) display the proximal LAD and mid RCA with reduced artifacts reconstructed with ECG-edit, as indicated by the white and black arrows, respectively. The ECG reports (H,N and I,O) with the algorithm-input HR identified as half of the original HR show that the reconstruction phases were 23% and 75%, respectively. BMI, body mass index; CAD, coronary artery disease; ECG, electrocardiogram; HR, heart rate; LAD, left anterior descending; RCA, right coronary artery.

CCTA objective image quality

The ECG-edited reconstructions demonstrated significant enhancements in the quantitative angiographic parameters, with the mean contrast density increasing from 215.82±88.37 to 290.37±82.34 HU (P<0.001), and the SNR improving from 5.41±3.27 to 6.99±3.96 (P=0.001). No significant difference was found between the two datasets in terms of noise (42.43±23.28 vs. 50.74±24.65, P=0.554). Table 4 provides detailed information on the objective image quality of the two datasets.

Discussion

To our knowledge, this study was the first to investigate the feasibility and the effect of this novel ECG-edit function, which deletes triggers alternately, in terms of its optimization of coronary artery visualization in CCTA in a clinical setting. In our study, the optimal phases, which primarily occurred at the 23% and 75% algorithm-input R-R interval, were largely selected from the data in which the ECG triggers were deleted from the first point, mainly in the end-systolic phase. Our results showed that the application of this novel ECG-edit function significantly improved subjective image quality while also reducing the number of non-interpretable coronary segments. The interpretability was higher at the per-segment, per-artery, and per-patient levels in the images with the ECG-edit than those without the ECG-edit. Superior objective image quality was achieved using the ECG-edit in the assessment of all studied coronary arteries as evidenced by the higher SNR values.

During retrospective helical scanning, to ensure that data for one cardiac cycle can be collected for every part of the heart in the Z-axis direction for later retrospective reconstruction, the table movement speed must be relatively slow and synchronized with the heart’s motion (38). The ECG-edit function is essential for arrhythmic cardiac MDCT, and a lower helical pitch should be set to avoid data deficits (24,25). In our study, a SnapShot Segment with a low helical pitch of 0.2 in patients with increased HR resulted in excessive redundant data acquisition, with approximately 80% overlap (39). This redundancy ensured sufficient data for reconstruction with the ECG-edit function when the interval triggers in normal rhythm were treated the same as those for arrhythmia, which supports the findings of Matsutani et al. (25). However, the mismatch between the increased HR and the low pitch, combined with the inherently faster coronary motion at a higher HR, resulted in inferior image quality, which was consistent with the lower image scores observed in the first dataset.

At a high HR (>75 bpm), previous studies (19,31) have indicated that the motion-free period in mid-diastole is reduced more significantly compared to that in end-systole, and the end-systolic phase reconstruction resulted in optimal image quality. In our study, the optimal reconstruction phase in the first study dataset was predominantly within the 40–59% R-R interval, which is consistent with prior research. In the second study dataset, the optimal reconstruction phases occurred primarily at 23% and 75% algorithm-input R-R interval. As illustrated in Figure 6, these phases corresponded to the systolic phase of the original cardiac cycle. This phenomenon occurs because the HR is calculated based on trigger points. After manually deleting trigger points at intervals, the algorithm-input HR becomes half of the actual true HR, and the cardiac cycle length doubles correspondingly. Thus, the 23% reconstruction phase (approximately 45%/2) corresponds to the end-systolic phase of the true cardiac cycle (approximately the 45% R-R interval) preceding the deleted trigger, while the 75% reconstruction phase (approximately 100%/2 + 45%/2) corresponds to the end-systolic phase of the true cardiac cycle following the deleted trigger point. This likely occurs because coronary artery motion is minimal during the end-systolic phase in patients with a high HR. Consequently, the optimal reconstruction phase with the least motion artifacts tends to mainly occur this period, regardless of whether the ECG-edit is performed.

Matsumoto et al. (22) and Matsutani et al. (24) reported that mid-diastolic images with the ECG-edit were of better quality than end-systolic images without the ECG-edit in patients with atrial fibrillation, due to the selective elimination of short R-R interval cycles. This protocol enabled optimized image synthesis from reduced cardiac motion data during slow HRs, strategically aligning the temporal resolution with the ventricular slow-filling phase. However, in our study, the optimal phases with the ECG-edit were mostly observed in the 23% and 75% algorithm-input R-R interval, which corresponds to the end-systolic phase of the cardiac cycle immediately preceding and following the deleted trigger, respectively. At a high HR (>65 bpm), the motion-free time in mid-diastole is reduced more significantly compared to that in end-systole (40,41). Unlike Matsutani et al. (24), who employed pharmacological HR modulation (target: 50–65 bpm via β-blocker administration), we enrolled patients with high and increased HRs, thereby optimizing cardiac phase shifts from mid-diastole to end-systole. However, further research is required to elucidate the specific mechanisms underlying the phenomenon observed in our results, where the optimal phase occurred more frequently at 23% than at 75%.

We also found that the optimal reconstruction phase was mostly selected from the data in which the ECG triggers were deleted starting from the first trigger. First, as illustrated in Figure 6C-6F, deletion starting from the first trigger reduced the algorithm-input HR to half of the actual HR while also decreasing HRV. Conversely, the ECG-edit method in which the deletion starts from the second trigger retained the first true cardiac cycle, resulting in greater HRV in the algorithm-input HR. Thus, HRV is a critical determinant of CCTA diagnostic fidelity, and high HRV results in variable end-diastolic cardiac volume and alters duration of the subsequent systolic phase (22). This limitation is particularly exacerbated in narrow-detector CT scanners that require multi-beat acquisition for complete cardiac imaging (13). This study included participants with HRV exceeding 30 bpm, but the ECG-edit function that deletes triggers alternately reduced the HRV, resulting in a significant improvement in the image quality of 64-slice CCTA. This finding aligns with previous research suggesting that HRV is significantly inversely correlated with both mean image quality and diagnostic accuracy in 64-slice CCTA (42). Second, as Figure 6C-6F show, when the ECG-edit function was used in which deletion started at the second trigger, not only was the number of complete reconstruction windows (as indicated by the blue vertical bars) available for reconstruction lower (4 in Figure 6E,6F vs. 5 in Figure 6C,6D), but the first or last original cardiac cycles retained also exhibited insufficient data during reconstruction (as indicated by the vertical bars with diagonal stripes in Figure 6E,6F).

In addition, our research also showed that images reconstructed with the ECG-edit function had higher SNR values. This improvement might be due to the function’s effectiveness in mitigating motion artifacts.

The ECG-edit strategy applied in this study serves as a post-processing solution. It cannot overcome the inherent physical temporal resolution of the CT scanner, which is constrained by the gantry rotation time; however, the ECG-edit function is widely available across various commercial systems, is not specific to certain manufacturers or models, and does not require additional software. These features highlight its potential for broad adoption in multicenter settings. We subsequently evaluated this method on an additional 64-slice CT scanner (Definition Edge, Siemens, Erlangen, Germany) at our institution. Using identical scanning parameters, we obtained images and employed the built-in ECG-edit tool to reconstruct images following the processing strategy developed in this study. The results remained consistent (see Figure S1 for a representative example), further affirming the technique’s clinical feasibility.

This study had several limitations. First, the generalizability of our findings is limited by the single-center retrospective design and relatively small sample size of the study, which may restrict its broader clinical applicability. All the data were acquired using a 64-slice MDCT scanner developed by a single manufacturer, which might have introduced bias into the current conclusions. Although consistent results were observed using another 64-slice CT scanner at our institution, further validation across multiple centers and scanner platforms is needed. Further, a substantial number of patients with prior coronary revascularization (stent placement or bypass grafting) were excluded from the study. Additionally, although several patients with elevated HRs (>80 bpm) were scanned using a reduced pitch after providing informed consent, and the proposed ECG-edit and reconstruction strategy still yielded high image quality, larger comparative studies are needed to evaluate image quality against high-pitch protocols in high-HR CCTA to confirm its broader applicability. Therefore, larger and more diverse cohorts, as well as prospective multicenter studies, need to be conducted to validate these results. Second, while this study assessed both subjective and objective image quality, the absence of automated phase-selection tools (e.g., SmartPhase) necessitated manual image evaluation at certain stages, introducing subjectivity and potentially affecting reproducibility. Further, the study did not evaluate coronary stenosis severity; thus, the effect of this ECG-edit on diagnostic accuracy for stenosis assessment was not explored. Nonetheless, the reduction in motion artifacts and the improvement in image quality suggest that the novel ECG-edit could improve the accuracy of detecting and quantifying coronary stenosis. Third, the diagnostic efficacy of CCTA was not compared with ICA as a clinical reference standard, although this was not the primary objective of the study. Fourth, due to limitations in the software configuration of the equipment, the ECG-edited images were not compared with the motion-corrected images, nor was the potential improvement in image quality investigated by combining the novel ECG-edit function with motion correction algorithms. Thus, future studies should conduct rigorous comparative evaluations of the image quality of edited coronary artery MDCT images with that of ICA and motion-corrected algorithm-processed images.

Conclusions

Our findings confirmed our hypothesis that alternate trigger deletion significantly enhances subjective image quality, interpretability, and quantitative image metrics in patients with increased HR. Further, the combination of alternating deletion starting from the first trigger and systolic phase reconstruction emerged as the optimal strategy for clinically acceptable images.

Acknowledgments

We express our gratitude to all individuals who have provided assistance and support throughout the course of this study. Their contributions have been invaluable and greatly appreciated.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-897/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-897/dss

Funding: This work was supported by the Scientific Research Project of Bishan Hospital of Chongqing, Bishan Hospital of Chongqing Medical University (grant No. BYKY2024013 to L.S.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-897/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Bishan Hospital of Chongqing Medical University (No. cqbykyll-20250110-2, dated January 10, 2025) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Collet JP, Thiele H, Barbato E, Barthélémy O, Bauersachs J, Bhatt DLESC Scientific Document Group, et al. 2020 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2021;42:1289-1367. [Crossref] [PubMed]
2. Serruys PW, Hara H, Garg S, Kawashima H, Nørgaard BL, Dweck MR, Bax JJ, Knuuti J, Nieman K, Leipsic JA, Mushtaq S, Andreini D, Onuma Y. Coronary Computed Tomographic Angiography for Complete Assessment of Coronary Artery Disease: JACC State-of-the-Art Review. J Am Coll Cardiol 2021;78:713-36. [Crossref] [PubMed]
3. Gulati M, Levy PD, Mukherjee D, Amsterdam E, Bhatt DL, Birtcher KK, et al. 2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021;144:e368-454. [Crossref] [PubMed]
4. van Rosendael SE, Shiyovich A, Cardoso RN, Souza Freire CV, van Rosendael AR, Lin FY, Larocca G, Bienstock SW, Blankstein R, Shaw LJ. The Role of Cardiac Computed Tomography Angiography in Risk Stratification for Coronary Artery Disease. J Soc Cardiovasc Angiogr Interv 2024;3:102230. [Crossref] [PubMed]
5. Abdelrahman KM, Chen MY, Dey AK, Virmani R, Finn AV, Khamis RY, Choi AD, Min JK, Williams MC, Buckler AJ, Taylor CA, Rogers C, Samady H, Antoniades C, Shaw LJ, Budoff MJ, Hoffmann U, Blankstein R, Narula J, Mehta NN. Coronary Computed Tomography Angiography From Clinical Uses to Emerging Technologies: JACC State-of-the-Art Review. J Am Coll Cardiol 2020;76:1226-43. [Crossref] [PubMed]
6. Dodd JD, Leipsic JA. Evolving Developments in Cardiac CT. Radiology 2023;307:e222827. [Crossref] [PubMed]
7. Soschynski M, Storelli R, Birkemeyer C, Hagar MT, Faby S, Schwemmer C, Nous FMA, Pugliese F, Vliegenthart R, Schlett CL, Nikolaou K, Krumm P, Nieman K, Bamberg F, Artzner CP. CT Myocardial Perfusion and CT-FFR versus Invasive FFR for Hemodynamic Relevance of Coronary Artery Disease. Radiology 2024;312:e233234. [Crossref] [PubMed]
8. Fan Y, Qin T, Sun Q, Wang M, Liang B. A Review of Factors Affecting Radiation Dose and Image Quality in Coronary CTA Performed with Wide-Detector CT. Tomography 2024;10:1730-43. [Crossref] [PubMed]
9. Mergen V, Sartoretti T, Baer-Beck M, Schmidt B, Petersilka M, Wildberger JE, Euler A, Eberhard M, Alkadhi H. Ultra-High-Resolution Coronary CT Angiography With Photon-Counting Detector CT: Feasibility and Image Characterization. Invest Radiol 2022;57:780-8. [Crossref] [PubMed]
10. Vecsey-Nagy M, Jermendy ÁL, Kolossváry M, Vattay B, Boussoussou M, Suhai FI, Panajotu A, Csőre J, Borzsák S, Fontanini DM, Csobay-Novák C, Merkely B, Maurovich-Horvat P, Szilveszter B. Heart Rate-Dependent Degree of Motion Artifacts in Coronary CT Angiography Acquired by a Novel Purpose-Built Cardiac CT Scanner. J Clin Med 2022;11:4336. [Crossref] [PubMed]
11. Miller RJH, Eisenberg E, Friedman J, Cheng V, Hayes S, Tamarappoo B, Thomson L, Berman DS. Impact of heart rate on coronary computed tomographic angiography interpretability with a third-generation dual-source scanner. Int J Cardiol 2019;295:42-7. [Crossref] [PubMed]
12. Chen Y, Liu Z, Li M, Yu Y, Jia Y, Ma G, Hu Z, Wei D, Li D, He T. Reducing both radiation and contrast doses in coronary CT angiography in lean patients on a 16-cm wide-detector CT using 70 kVp and ASiR-V algorithm, in comparison with the conventional 100-kVp protocol. Eur Radiol 2019;29:3036-43. [Crossref] [PubMed]
13. Liang J, Sun Y, Ye Z, Sun Y, Xu L, Zhou Z, Thomsen B, Li J, Sun Z, Fan Z. Second-generation motion correction algorithm improves diagnostic accuracy of single-beat coronary CT angiography in patients with increased heart rate. Eur Radiol 2019;29:4215-27. [Crossref] [PubMed]
14. Lossau Née Elss T, Nickisch H, Wissel T, Bippus R, Schmitt H, Morlock M, Grass M. Motion estimation and correction in cardiac CT angiography images using convolutional neural networks. Comput Med Imaging Graph 2019;76:101640. [Crossref] [PubMed]
15. Liu S, Huang L, Gui S, Pang X, Liu J, Li X, Wang Y, He W, Zhang X, Peng L. Optimizing coronary CT angiography quality with motion-compensated reconstruction for second-generation dual-layer spectral detector CT. Eur Radiol 2025;35:381-92. [Crossref] [PubMed]
16. Yamaguchi S, Ichikawa Y, Takafuji M, Sakuma H, Kitagawa K. Usefulness of second-generation motion correction algorithm in improving delineation and reducing motion artifact of coronary computed tomography angiography. J Cardiovasc Comput Tomogr 2024;18:281-90. [Crossref] [PubMed]
17. Bastarrika G, Lee YS, Huda W, Ruzsics B, Costello P, Schoepf UJ. CT of coronary artery disease. Radiology 2009;253:317-38. [Crossref] [PubMed]
18. Carrascosa P, Capuñay C, Deviggiano A, Goldsmit A, Tajer C, Bettinotti M, Carrascosa J, Ivanc TB, Fallahi A, García MJ. Accuracy of low-dose prospectively gated axial coronary CT angiography for the assessment of coronary artery stenosis in patients with stable heart rate. J Cardiovasc Comput Tomogr 2010;4:197-205. [Crossref] [PubMed]
19. Yang CC, Law WY, Lu KM, Wu TH. Relationship between heart rate and optimal reconstruction phase in coronary CT angiography performed on a 256-slice multidetector CT. Br J Radiol 2019;92:20180945. [Crossref] [PubMed]
20. Zhang J, Fletcher JG, Scott Harmsen W, Araoz PA, Williamson EE, Primak AN, McCollough CH. Analysis of heart rate and heart rate variation during cardiac CT examinations. Acad Radiol 2008;15:40-8. [Crossref] [PubMed]
21. Ghekiere O, Salgado R, Buls N, Leiner T, Mancini I, Vanhoenacker P, Dendale P, Nchimi A. Image quality in coronary CT angiography: challenges and technical solutions. Br J Radiol 2017;90:20160567. [Crossref] [PubMed]
22. Matsumoto H, Kondo T, Watanabe S, Kikumoto R, Shimada T, Hiraoka Y, Ueda K. ECG-edited middiastolic phase reconstruction improves image quality at 64-MDCT coronary angiography of patients with atrial fibrillation. AJR Am J Roentgenol 2008;191:1659-66. [Crossref] [PubMed]
23. Cho YJ, Choe YH, Lee MS. Comparison of image quality of 64-slice multidetector CT coronary CT angiography using automated and manual multiphase methods for the determination of optimal phases for image reconstruction in patients with various mean heart rates. Int J Cardiovasc Imaging 2010;26:41-52. [Crossref] [PubMed]
24. Matsutani H, Sano T, Kondo T, Morita H, Arai T, Sekine T, Takase S, Oida A, Fukazawa H, Suguta M, Kondo M, Kodama T, Orihara T, Yamada N, Tsuyuki M, Narula J. ECG-edit function in multidetector-row computed tomography coronary arteriography for patients with arrhythmias. Circ J 2008;72:1071-8. [Crossref] [PubMed]
25. Matsutani H, Sano T, Kondo T, Sekine T, Arai T, Morita H, Takase S. Optimization of helical pitch in cardiac MDCT acquisition in patients with arrhythmia requiring ECG-edit. Nihon Hoshasen Gijutsu Gakkai Zasshi 2008;64:1343-51. [Crossref] [PubMed]
26. Sano T, Matsutani H, Kondo T, Fujimoto S, Sekine T, Arai T, Morita H, Takase S. Estimation of radiation dose and image quality of coronary 320-row area detector CT angiography by optimal prospective ECG-gated protocols for different heart rate. Nihon Hoshasen Gijutsu Gakkai Zasshi 2011;67:1398-407. [Crossref] [PubMed]
27. Matsutani H, Sano T, Kondo T, Fujimoto S, Sekine T, Arai T, Morita H, Takase S. Reconstruction of coronary 64-multidetector-row computed tomography angiography in patients with atrial fibrillation: delete short RR intervals using electorocardiogram. Nihon Hoshasen Gijutsu Gakkai Zasshi 2012;68:50-8. [Crossref] [PubMed]
28. Sabarudin A, Sun Z, Ng KH. Coronary computed tomography angiography with prospective electrocardiography triggering: a systematic review of image quality and radiation dose. Singapore Med J 2013;54:15-23. [Crossref] [PubMed]
29. Yang L, Xu L, Schoepf UJ, Wichmann JL, Fox MA, Yan J, Fan Z, Zhang Z. Prospectively ECG-Triggered Sequential Dual-Source Coronary CT Angiography in Patients with Atrial Fibrillation: Influence of Heart Rate on Image Quality and Evaluation of Diagnostic Accuracy. PLoS One 2015;10:e0134194. [Crossref] [PubMed]
30. Kaniewska M, Deininger-Czermak E, Getzmann JM, Wang X, Lohezic M, Guggenberger R. Application of deep learning-based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time. Eur Radiol 2023;33:1513-25. [Crossref] [PubMed]
31. Seifarth H, Wienbeck S, Püsken M, Juergens KU, Maintz D, Vahlhaus C, Heindel W, Fischbach R. Optimal systolic and diastolic reconstruction windows for coronary CT angiography using dual-source CT. AJR Am J Roentgenol 2007;189:1317-23. [Crossref] [PubMed]
32. Celeng C, Vadvala H, Puchner S, Pursnani A, Sharma U, Kovacs A, Maurovich-Horvat P, Hoffmann U, Ghoshhajra B. Defining the optimal systolic phase targets using absolute delay time for reconstructions in dual-source coronary CT angiography. Int J Cardiovasc Imaging 2016;32:91-100. [Crossref] [PubMed]
33. Balaney B, Vembar M, Grass M, Singh A, Kawaji K, Landeras L, Chung J, Mor-Avi V, Patel AR. Improved visualization of the coronary arteries using motion correction during vasodilator stress CT myocardial perfusion imaging. Eur J Radiol 2019;114:1-5. [Crossref] [PubMed]
34. Pontone G, Bertella E, Mushtaq S, Loguercio M, Cortinovis S, Baggiano A, Conte E, Annoni A, Formenti A, Beltrama V, Guaricci AI, Andreini D. Coronary artery disease: diagnostic accuracy of CT coronary angiography--a comparison of high and standard spatial resolution scanning. Radiology 2014;271:688-94. [Crossref] [PubMed]
35. Mergen V, Sartoretti T, Cundari G, Serifovic M, Higashigaito K, Allmendinger T, Schmidt B, Flohr T, Manka R, Eberhard M, Alkadhi H. The Importance of Temporal Resolution for Ultra-High-Resolution Coronary Angiography: Evidence From Photon-Counting Detector CT. Invest Radiol 2023;58:767-74. [Crossref] [PubMed]
36. Halliburton SS, Abbara S, Chen MY, Gentry R, Mahesh M, Raff GL, Shaw LJ, Hausleiter JSociety of Cardiovascular Computed Tomography. SCCT guidelines on radiation dose and dose-optimization strategies in cardiovascular CT. J Cardiovasc Comput Tomogr 2011;5:198-224. [Crossref] [PubMed]
37. Pontone G, Andreini D, Bertella E, Baggiano A, Mushtaq S, Loguercio M, Segurini C, Conte E, Beltrama V, Annoni A, Formenti A, Petullà M, Guaricci AI, Montorsi P, Trabattoni D, Bartorelli AL, Pepi M. Impact of an intra-cycle motion correction algorithm on overall evaluability and diagnostic accuracy of computed tomography coronary angiography. Eur Radiol 2016;26:147-56. [Crossref] [PubMed]
38. Halliburton S, Arbab-Zadeh A, Dey D, Einstein AJ, Gentry R, George RT, Gerber T, Mahesh M, Weigold WG. State-of-the-art in CT hardware and scan modes for cardiovascular CT. J Cardiovasc Comput Tomogr 2012;6:154-63. [Crossref] [PubMed]
39. Feldle P, Grunz JP, Huflage H, Kunz AS, Ergün S, Afat S, Gruschwitz P, Görtz L, Pennig L, Bley TA, Conrads N. Influence of helical pitch and gantry rotation time on image quality and file size in ultrahigh-resolution photon-counting detector CT. Sci Rep 2024;14:9358. [Crossref] [PubMed]
40. Achenbach S, Ropers D, Holle J, Muschiol G, Daniel WG, Moshage W. In-plane coronary arterial motion velocity: measurement with electron-beam CT. Radiology 2000;216:457-63. [Crossref] [PubMed]
41. Herzog C, Abolmaali N, Balzer JO, Baunach S, Ackermann H, Dogan S, Britten MB, Vogl TJ. Heart-rate-adapted image reconstruction in multidetector-row cardiac CT: influence of physiological and technical prerequisite on image quality. Eur Radiol 2002;12:2670-8. [Crossref] [PubMed]
42. Leschka S, Scheffel H, Husmann L, Gämperli O, Marincek B, Kaufmann PA, Alkadhi H. Effect of decrease in heart rate variability on the diagnostic accuracy of 64-MDCT coronary angiography. AJR Am J Roentgenol 2008;190:1583-90. [Crossref] [PubMed]