Assessment of atrial septal defect size during systole and diastole using 4D CT: a case series
Yukihiro Nishimoto
, Takumi Kawase, Ryo Fujii, Ryo Nangoya, Munehide Nagao, Kazuki Noda, Goki Inno, Kenta Nishiya, Yosuke Takahashi
Contributions: (I) Conception and design: Y Nishimoto; (II) Administrative support: None; (III) Provision of study materials or patients: All authors; (IV) Collection and assembly of data: Y Nishimoto; (V) Data analysis and interpretation: Y Nishimoto; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.
Background: The size of atrial septal defects (ASDs) is typically measured using echocardiography. In contrast, computed tomography (CT) images offer superior spatial resolution, and four-dimensional (4D) CT enables visualization of the defect hole during various cardiac phases. In this study, we compared the results of 4D CT measurements of defect sizes during atrial systole (As) and atrial diastole (Ad) in 7 of 18 cases of robot-assisted ASD closure (Robotic ASD closure) with measurements obtained by other methods.
Case Description: Robotic ASD closure was performed in 18 patients between December 2019 and April 2025. Surgery was performed under general anesthesia, with a venous line from the right internal jugular vein and right femoral vein, and an arterial line on the right femoral artery to establish extracorporeal circulation. The procedure was conducted using the DaVinci Xi® surgical system. Image analysis was performed using VINCENT®, and the size of the defect was measured during As and Ad, and calculated using the formula: long radius × short radius × π. The mean age was 60.9 years (range, 34–82 years), and 10 (55.6%) patients were male. No residual shunt flow was detected on intraoperative or postoperative echocardiography. The defect size was measurable using 4D CT in 7 cases. The area of the defect was significantly different between As [65.9–207.2 mm2, mean 134.1 mm2, interquartile range (IQR) 58.9] and Ad (113.0–490.6 mm2, mean 226.1 mm2, IQR 94.2) (P=0.04).
Conclusions: Using 4D CT, it was possible to determine the morphology and area of ASD defects in As/Ad before surgery. This is useful for selecting surgical procedures and determining patch size and may contribute to reducing cardiac arrest and surgical time. To prove that 4D CT can accurately predict the size and morphology of the defects in As/Ad, further accumulation of cases and analyses are required.
Keywords: Robotic surgery; atrial septal defect (ASD); four-dimensional computed tomography (4D CT); case series
Received: 31 August 2025; Accepted: 27 October 2025; Published online: 08 January 2026.
doi: 10.21037/acr-2025-231
Highlight box
Key findings
• Atrial systole and diastole were estimated to be approximately 5% and 45% respectively of the RR interval timing in four-dimensional computed tomography (4D CT).
• The imaging protocol will use the current one, and image analysis will be performed at 0% and 40%.
• The area of the defect contracted significantly and expanded between atrial systole and diastole.
What is known and what is new?
• The defect size is typically measured using echocardiography. On the other hand, CT offers superior spatial resolution and enables permanent storage, reconstruction, and analysis. Furthermore, 4D CT allows measurement of the defect size at various points in the cardiac cycle.
What is the implication, and what should change now?
• By utilizing 4D CT, the size and morphology of the defect in systole and diastole can be assessed preoperatively, providing a reference for selecting between direct and patch closures. The size and shape of the patch can be predicted may allowing the patch to be prepared in advance at the time of pericardial incision when using autologous pericardium, or prior to cardiac arrest when using polyester or bovine pericardium. These measures may contribute to reductions in cardiac arrest and surgical time.
• At this stage, 4D CT can serve as a tool to more robustly support evaluations obtained through conventional echocardiography.
Introduction
Background
Atrial septal defect (ASD) is the most common congenital heart disease encountered in adulthood, excluding bicuspid aortic valve disease, accounting for 35–40% of congenital heart diseases in individuals over 40 years of age (1). Depending on the location of the defect, it is classified into secundum, partial atrioventricular defect (pAVSD), sinus venosus defect (SVD), and coronary sinus defect (CSD) types; however, SVD and CSD types are rare. It is reported to be more common in females at a ratio of approximately 2:1 and generally, surgical repair is considered when Qp/Qs >1.5 (2).
Rationale and knowledge gap
The defect size is typically measured using echocardiography. On the other hand, computed tomography (CT) offers superior spatial resolution and enables permanent storage, reconstruction, and analysis. Furthermore, four-dimensional computed tomography (4D CT) allows measurement of the defect size at various points in the cardiac cycle.
Objective
In this study, we evaluated the utility of 4D CT by measuring the size of the defect during atrial systole (AS) and atrial diastole (Ad) in 18 cases of robotic ASD closure and compared the results with those obtained from other imaging modalities. We present this article in accordance with the PROCESS and AME Case Series reporting checklists (available at https://acr.amegroups.com/article/view/10.21037/acr-2025-231/rc).
Case presentation
Patients
This study is a single-center retrospective and consecutive case series. Robotic ASD closure was performed in 18 patients between December 2019 and April 2025. There were no cases of ASD alone. All patients underwent concomitant procedures, including mitral valve surgery, tricuspid valve surgery, left atrial appendage closure (LAAC), or the Maze procedure, either alone or in combination. This study was registered in the Research Registry (researchregistry11496). All procedures performed in this article were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patients for the publication of this article and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.
Surgical procedure
After the patient was placed under general anesthesia, the anesthesiologist and cardiac surgeon inserted a 19 Fr vascular access device into the right internal jugular vein via puncture, advancing it immediately before the junction of the right brachiocephalic vein. The patient was placed in the left lateral recumbent position, and a 4–5 cm incision was made in the fourth intercostal space, which was designated as the main incision site. The ports were inserted into the third, fourth, and sixth intercostal spaces. A cannula was inserted into the right internal jugular vein via the femoral vein under fluoroscopic guidance until it overlapped with the existing cannula by 1 cm. An atrial line was inserted into the right femoral artery, and a cardiopulmonary bypass was established. The robot was used in all cases with the DaVinci Xi® surgical system (Intuitive Surgical Inc, Sunnyvale, CA, USA). The DaVinci® was rolled in, ports were placed at the main incision site, the camera was equipped, and targeting was performed. Ports were placed at the third, fourth, and sixth intercostal spaces, with the left hand (DeBakey forceps®), Dual blade retractor®, and right hand (Permanent cautery spatula®, Large suture cut needle driver®, or Monopolar curved scissors®) respectively. Aortic cross-clamping and assistant surgeon maneuvers were performed at the primary incision site.
Image acquisition and measurement
In our facility’s cardiac CT protocol, the RR interval is set to 100%, and 128 or 160 slices are acquired in 10 phases at 10% intervals from the previous R-wave. Additionally, 256 and 320 slices were acquired in approximately 40% and 75% of the ventricular systolic (Vs) and diastolic (Vd) phases, respectively. For 3D image reconstruction, we used the image analysis software VINCENT® (Fujifilm®, Tokyo, Japan). For the size of the defect during As and Ad, the measurements at phase 0% of 4D CT were used for As and intraoperative measurements were used for all cases of Ad. The area of the defect was assumed to be elliptical, and the area of the ellipse was calculated using the formula: major radius × minor radius × π.
Statistical analysis
The areas of As and Ad, which do not follow a normal distribution, were compared between two paired groups from the same patient using the Wilcoxon signed-rank test. Statistical analyses were performed using EZR version 4.3.1.
Results
The mean patient age was 60.9 years (range, 34–82 years), and 10 (55.6 %) patients were male. All cases were closed using autologous pericardium, bovine pericardium, or polyester patches, and no direct closure was performed. No residual shunt flow was detected on intra-or postoperative echocardiography. Concomitant procedures included tricuspid annuloplasty (TAP) in 16 patients, mitral annuloplasty (MAP) in 3 patients, mitral valvuloplasty (MVP) with MAP in 2 patients, mitral valve replacement (MVR) in 1 patient, left atrial plication (LAP) in 1 patient, AF-related procedures, LAAC in 4 patients, and maze surgery in 1 patient. Tricuspid valvuloplasty (TVP) without TAP was performed in a single case of pAVSD. Two patients had four or more defects, one had three defects, two had two defects, and the remaining 13 had single-hole defects. The preoperative rhythm was AF in three cases and sinus rhythm in the remaining 15 cases. The Qp/Qs ratio immediately before surgery was 1.8–4.2 (mean, 2.4). 1 patient had pAVSD, 1 had an inferior sinus venosus defect (ISVD), 1 had a CSD, and the remaining 15 had a secondary defect (Table 1).
Table 1
| Patient | Age, years | Gender | Concomitant procedure | Pore | Rhythm | Qp/Qs | Remarks | Exclusion reasons |
|---|---|---|---|---|---|---|---|---|
| 1 | 52 | F | TAP | 1 | SR | 2.5 | ||
| 2 | 75 | F | MVR, TAP, LAAC, LAP | 1 | AF | 4.2 | Cannot depict | |
| 3 | 66 | M | MVP, TAP | 3 | SR | 1.8 | Cannot depict | |
| 4 | 79 | F | TAP | 1 | SR | 2.4 | ||
| 5 | 37 | F | TAP | 1 | SR | 2.0 | ||
| 6 | 34 | M | TAP | 1 | SR | 1.8 | 3D CT only | 3D CT only |
| 7 | 82 | M | MAP, TAP, LAAC | Multiple | AF | 2.4 | ISVD | Cannot depict |
| 8 | 68 | M | TAP | 1 | SR | 1.8 | ||
| 9 | 72 | M | TAP | 1 | SR | 2.1 | CSD | |
| 10 | 61 | M | TAP | Multiple | SR | N/A | Cannot depict | |
| 11 | 59 | M | TAP | 1 | SR | 2.4 | 3D CT only | 3D CT only |
| 12 | 67 | M | TAP | 1 | SR | 3.0 | No cardiac CT | No cardiac CT |
| 13 | 66 | F | MAP, TAP | 2 | SR | 3.6 | No cardiac CT | No cardiac CT |
| 14 | 45 | F | MVP | 1 | SR | N/A | ||
| 15 | 59 | F | TVP | 1 | SR | 2.0 | pAVSD | |
| 16 | 51 | M | MAP, TAP, Maze, LAAC | 1 | AF | 2.1 | No cardiac CT | No cardiac CT |
| 17 | 73 | M | TAP, LAAC | 1 | SR | 1.8 | 3D CT only | 3D CT only |
| 18 | 50 | F | TAP | 2 | SR | 2.1 | No cardiac CT | No cardiac CT |
The average age was 60.9 years, with 10 cases (55.6%) being male. Qp/Qs ranged from 1.8 to 4.2 (average 2.4). 3D, three-dimensional; AF, atrial fibrillation; CSD, coronary sinus defect; CT, computed tomography; F, female; ISVD, inferior sinus venosus defects; LAAC, left atrial appendage closure; LAP, left arterial plication; M, male; MAP, mitral annuloplasty; MVP, mitral valvuloplasty; MVR, mitral valve replacement; N/A, not available; pAVSD, partial atrioventricular septum defect; SR, sinus rhythm; TAP, tricuspid annuloplasty; TVP, tricuspid valvuloplasty.
In the image analysis, 7 patients were excluded: 3 in which 3D cardiac CT was performed in a single phase, and 4 in which cardiac CT was not performed. Among the remaining 11 patients, 4 cases in which the outline of the defect was not clearly visualized were excluded, and the final analysis included 7 cases (Table 2).
Table 2
| Patient | CT As (mm) | CT As area (mm2) | CT Ad (mm) | CT Ad area (mm2) | ECG (mm) | ECG area (mm2) | IO (mm) | IO area (mm2) | Remarks |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 11×24 | 207.2 | N/A | N/A | 17×20 | 266.9 | 15×20 | 235.5 | |
| 4 | 10×15 | 117.8 | N/A | N/A | 16×19 | 238.6 | 16×19 | 213.5 | |
| 5 | 9×15 | 106.0 | 8×18 | 113.0 | 12×15 | 141.3 | 12×15 | 141.3 | |
| 8 | 6×14 | 65.9 | 9×17 | 120.1 | 9×14 | 98.9 | 13×15 | 153.1 | |
| 9 | 14×15 | 164.9 | 13×16 | 163.3 | N/A | N/A | 25×25 | 490.6 | CSD |
| 14 | 12×12 | 113.0 | N/A | N/A | 12×15 | 141.3 | 12×12 | 113.0 | |
| 15 | 11×19 | 164.1 | 14×19 | 208.8 | N/A | N/A | 15×20 | 235.5 | pAVSD |
The area of the defect was calculated using CT 0% as the atrial contraction phase and intraoperative measurements as the atrial diastole phase. Ad, atrial diastole; As, atrial systole; CSD, coronary sinus defect; CT, computed tomography; ECG, echocardiography; IO, intraoperative; N/A, not available; pAVSD, partial atrioventricular septum defect.
The area of the defect was measured as As [65.9–207.2 mm2, mean 134.1 mm2, interquartile range (IQR) 58.9] compared to Ad (113.0–490.6 mm2, mean 226.1 mm2, IQR 94.2), with a statistically significant difference (P=0.04) (Figure 1).
Discussion
As the first key finding, As and Ad were estimated to be approximately 5% and 45% of the RR interval timing in 4D CT, respectively. In this study, we first considered which of the 10 phases of 4D CT corresponded to As and Ad. In the electrocardiogram, atrial depolarization occurs during the P wave, and repolarization occurs during the QRS complex; therefore, As is considered to occur immediately after the P wave and Ad immediately after the QRS complex. However, it is unclear to which phases As and Ad correspond. Therefore, we observed the time points at which the atria were most dilated and contracted in each phase. The phase with the highest contraction was at 0–10%, and the phase with the highest dilation was at 40–50%. This was consistent across all cases in where 4D CT imaging was possible. Second, the current imaging protocol, which uses thin slices only at approximately 40% and 75%, is considered sufficient for preoperative CT in ASD closure. In our facility’s 4D CT, only the 40% phase corresponding to the Vs and the 75% phase corresponding to the Vd were imaged with 256 or 320 slices, whereas the remaining 10 phases were imaged with 128 or 160 slices, respectively. As both Ad and Vs were found to be approximately 40%, detailed images of Vs were used to measure the defect size in Ad. Additionally, the defect size can be measured at 0% using nonthin slices in As. In summary, we propose using the current imaging protocol, analyzing As using 0% phase normal slice and Ad using 40% phase thin slices. Third, since the defect region was found to vary significantly between As and Ad, the closure method must be selected considering size and morphology.
This study has limitations. First, under the current imaging protocol, no cases of defect visualization were achieved in patients with atrial fibrillation (AF) or multi-hole ASD, rendering it applicable only to patients with single-hole sinus rhythm. Development of an imaging protocol capable of visualizing defects in patients with AF or multi-hole ASD is necessary. Second, the defect area was calculated assuming an elliptical shape. It is not possible to accurately calculate the area, particularly in cases where the defect has a complex shape, necessitating more precise area measurement techniques such as manual contour tracing. Third, this study was limited to robotic ASD closure at a single institution and did not include cases of sternal median incision or thoracoscopic procedures. Additionally, patients who did not undergo 4D CT imaging were excluded. Further validation with a larger number of cases is necessary to prove that 4D CT can accurately predict the size and morphology of defects in As/Ad.
Reports utilizing 3D printing for preoperative planning and technique confirmation in ISVD catheterization have been reported (3,4). He et al. (3) constructed a 3D CT at a phase of 75% (Vd) with 256 slices; however, no reports focusing on the timing of As/Ad or mentioning changes in defect size and morphology were found. A summary of the midterm outcomes of thoracoscopic or robotic surgery for ASD (5-7) indicates that direct closure was selected when the defect was small and excessive tension was not expected, resulting in direct closure in 88 cases (32.5%) and patch closure in 183 cases (67.5%). Except for one case in which the patch detached (7), there were no cases of residual shunt flow or reoperation. This suggests that surgeons can accurately select between direct and patch closure based on their experience.
In the seven cases presented here, the size of the ASD was measurable using CT in only four cases; however, these results were comparable to those obtained using echocardiography or intraoperative measurements. Compared with As, Ad provided images that were twice as detailed; however, in cases where measurement was impossible, the atrial septum and atrial cavity could not be distinguished, and the defect on the image appeared to be of an unrealistic size. This is thought to be because the atrial septum becomes thinner in Ad, making it difficult to distinguish between the septum and cavity. Therefore, because the atrial myocardium relaxes during cardiac arrest, it was assumed that its size would be similar to that of the Ad, and intraoperative measurement results were used. However, the intraoperative measurement results cannot be definitively used to represent the true defect hole size of the Ad, as blood pressure is absent during cardiac arrest. To depict the defect clearly, it is desirable to acquire thin slices in all phases; however, this would unnecessarily increase the patient’s radiation exposure. In this study, no visualizable cases were observed in patients with multi-hole ASD or AF. This may be because in cases with multiple defects, the spaces between the defects are often membrane-like, and the shunt flow from multiple defects causes the contrast agent to be distributed unevenly in the atrial cavity. In AF, atrial contractions are irregular, causing an uneven distribution of the contrast agent and irregular RR intervals. Further research is needed to develop imaging protocols capable of visualizing in patients with multi-hole ASD or AF, and to identify specific factors contributing to imaging difficulties. In conclusion, no clear correlation was observed between Qp/Qs and defect hole size, and the current imaging protocol was considered effective only in single-hole ASD and sinus rhythm cases. In Patient 9, there was little change in the defect hole size between As and Ad, and the intraoperatively measured values were significantly different from those in other cases. This is thought to be due to the CSD, where the defect was membrane-like, resulting in capture of the orifice of the coronary sinus.
This methods requires invasive testing using contrast-enhanced CT. However, contrast-enhanced CT is indispensable as a preoperative test for cardiac surgery, except in cases of severe renal impairment or severe contrast allergy. Therefore, no additional invasive procedures are necessary for this purpose. Utilizing 4D CT allows preoperative assessment of the size and morphology of the As/Ad defect, enabling selection criteria for direct closure versus patch closure. Furthermore, sharing changes in defect size and morphology with other surgeons facilitates preoperative case discussions within the cardiac team and streamlines surgeon training. Additionally, when performing patch closure, the size and shape of the patch can be predicted. This allows for advance preparation of the patch: during pericardial incision when using autologous pericardium, or before cardiopulmonary bypass when using polyester or bovine pericardium. These measures have the potential to contribute to reduced cardiopulmonary bypass time and shorter overall operative time. This study found that the accuracy of defect hole assessment using 4D CT was still not satisfactory. Further research is needed to improve the reliability of 4D CT evaluation. At this stage, 4D CT can serve as a tool to reinforce assessments made with conventional echocardiography.
Conclusions
Using 4D CT, it is possible to pre-emptively assess changes in the morphology and area of ASD defects in the As/Ad region, which can serve as a reference for surgical procedure selection and patch size determination, potentially contributing to reduced cardiac arrest and surgical procedure times.
Further investigation with a larger number of cases is necessary to verify the accurate predictive capability of 4D CT for assessing the morphology and size of defects in the As/Ad region.
Acknowledgments
The contents of this manuscript were presented at the Kansai Thoracic Surgical Association 2025 (Tsu, Mie, Japan).
Footnote
Reporting Checklist: The authors have completed the PROCESS and AME Case Series reporting checklists. Available at https://acr.amegroups.com/article/view/10.21037/acr-2025-231/rc
Peer Review File: Available at https://acr.amegroups.com/article/view/10.21037/acr-2025-231/prf
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://acr.amegroups.com/article/view/10.21037/acr-2025-231/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. This study was registered in the Research Registry (researchregistry11496). All procedures performed in this article were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patients for the publication of this article and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.
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
- Geva T, Martins JD, Wald RM. Atrial septal defects. Lancet 2014;383:1921-32. [Crossref] [PubMed]
- JCS/JCC/JSCS/JSVS/JATS 2021 Guideline on catheter intervention for congenital heart disease and structural heart disease. Available online: https://www.j-circ.or.jp/cms/wp-content/uploads/2021/03/JCS2021_Sakamoto_Kawamura.pdf
- He L, Cheng GS, Du YJ, et al. Feasibility of Device Closure for Multiple Atrial Septal Defects With an Inferior Sinus Venosus Defect: Procedural Planning Using Three-Dimensional Printed Models. Heart Lung Circ 2020;29:914-20. [Crossref] [PubMed]
- Thakkar AN, Chinnadurai P, Breinholt JP, et al. Transcatheter closure of a sinus venosus atrial septal defect using 3D printing and image fusion guidance. Catheter Cardiovasc Interv 2018;92:353-7. [Crossref] [PubMed]
- Liu G, Qiao Y, Zou C, et al. Totally thoracoscopic surgical treatment for atrial septal defect: mid-term follow-up results in 45 consecutive patients. Heart Lung Circ 2013;22:88-91. [Crossref] [PubMed]
- Xiao C, Gao C, Yang M, et al. Totally robotic atrial septal defect closure: 7-year single-institution experience and follow-up. Interact Cardiovasc Thorac Surg 2014;19:933-7. [Crossref] [PubMed]
- Lee H, Yang JH, Jun TG, et al. The Mid-term Results of Thoracoscopic Closure of Atrial Septal Defects. Korean Circ J 2017;47:769-75. [Crossref] [PubMed]
Cite this article as: Nishimoto Y, Kawase T, Fujii R, Nangoya R, Nagao M, Noda K, Inno G, Nishiya K, Takahashi Y. Assessment of atrial septal defect size during systole and diastole using 4D CT: a case series. AME Case Rep 2026;10:49.
