Dual-volume reconstruction technology in the detection and classification of recurrent aneurysms following intracranial aneurysm embolization

Introduction

Intracranial aneurysms (ICAs) represent a prevalent category of neurovascular disorders in clinical practice, with their rupture frequently leading to subarachnoid hemorrhage (SAH), a condition associated with substantial mortality and morbidity rates (1,2). From a pathophysiological perspective, hemodynamic stress-induced vessel wall weakening and subsequent degenerative changes constitute key drivers of aneurysm initiation and rupture (3). Notably, additional hemodynamic studies reveal that aneurysm wall enhancement (AWE) correlates with areas of low wall shear stress, a hemodynamic pattern linked to proinflammatory changes and wall instability (4-6). Advances in interventional neuroradiology have positioned embolization as a frontline treatment option for these aneurysms. Nevertheless, recurrence following embolization can give rise to recurrent aneurysms, which elevate the risk of re-rupture and may necessitate additional interventional measures (7,8). Consequently, the accurate detection and classification of recurrent aneurysms in the postoperative period, achieved through dual-volume reconstruction technology (DVRT)’s high diagnostic accuracy, are crucial for reducing complications and improving patient outcomes. By reliably identifying even small, irregularly shaped, and posterior circulation aneurysms, DVRT allows clinicians to more precisely stratify patient risk, thereby directly guiding personalized treatment planning.

Currently, a variety of imaging techniques, including computed tomography angiography (CTA), magnetic resonance angiography (MRA), and two-dimensional digital subtraction angiography (2D-DSA), are routinely utilized for postoperative surveillance of ICA. Among these, 2D-DSA is regarded as the gold standard due to its exceptional sensitivity and specificity (9-12). However, its invasive nature and associated complication risks restrict its frequent use in routine follow-up. CTA and MRA, as non-invasive alternatives, exhibit suboptimal detection capabilities for recurrent aneurysms, particularly those that are small or morphologically complex (11,13,14).

In this context, DVRT emerges as a promising imaging modality. Leveraging multi-angle scanning and high-resolution reconstruction, DVRT offers unparalleled visualization of postoperative recurrent aneurysms. It distinctly delineates aneurysm boundaries and captures intricate morphological details, facilitating enhanced detection rates due to its ease of operation and superior image resolution (15,16). Recent research has demonstrated considerable promise for DVRT in the postoperative surveillance of patients with ICA (17,18). However, rigorous validation is still needed to establish its specific utility in detecting and classifying recurrent aneurysms. This study was designed to compare the diagnostic efficacy of DVRT versus CTA and MRA in detecting aneurysm recurrence following ICA embolization, while elucidating its clinical utility. Furthermore, we examined the correlations between the morphology, size, and location of recurrent aneurysms and post-embolization clinical outcomes. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-303/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Zhongda Hospital Affiliated to Southeast University (No. 2020ZDSYLL158-P01) and the requirement for individual consent for this retrospective analysis was waived.

Patients selection and follow-up protocol

This study was a retrospective cohort study including patients who underwent ICA embolization and postoperative follow-up in the Center of Interventional Radiology and Vascular Surgery at Zhongda Hospital Affiliated to Southeast University between January 2020 and December 2023. At the Center of Interventional Radiology and Vascular Surgery at Zhongda Hospital Affiliated to Southeast University, DVRT is integrated into standard postoperative surveillance protocols to complement CTA and MRA, addressing limitations in spatial resolution and contrast sensitivity for small or morphologically complex aneurysms. All enrolled patients had undergone 2D-DSA within 48 hours after baseline CTA and MRA examinations during follow-up for confirmation of aneurysm recurrence. The projection angles for 2D-DSA were determined using three-dimensional digital subtraction angiography (3D-DSA) to obtain tangential views of the aneurysm neck. These angles were selected to optimize the visualization of recurrent aneurysm neck morphology patterns. Recurrent aneurysms identified by CTA, MRA, and DVRT were validated against 2D-DSA results; recurrence assessments were independently conducted by two neurointerventional radiologists with over 15 years of experience who were blinded to baseline imaging findings, with discordant cases resolved through adjudication by a third senior neurointerventional radiologist. The inclusion criteria were as follows: (I) diagnosis of ICA and treatment with endovascular embolization; (II) postoperative follow-up for at least 12 months with complete imaging data (including CTA, MRA, and DVRT imaging); and (III) aged between 18 and 85 years. The exclusion criteria were as follows: (I) severe neurological damage or other major diseases affecting prognosis before or after surgery; (II) incomplete follow-up or imaging data after surgery or residual aneurysms immediately post-embolization; and (III) unsuitable for participation in the study for other reasons.

To comply with clinical guidelines requiring a ≥12-month surveillance period for detecting recurrence, all patients underwent scheduled follow-up imaging (CTA, MRA, and DVRT) and clinical assessments at 3 and 12 months postoperatively, ensuring both timely detection and longitudinal outcome analysis (19,20).

Image acquisition and image preprocessing

This study aimed to evaluate the clinical utility of DVRT by comparing its performance with traditional imaging methods (CTA and MRA) in detecting postoperative recurrent aneurysms. The study consisted of the following steps.

Imaging data acquisition

Preoperative and postoperative imaging evaluations for all patients followed standardized protocols (CTA, MRA, and DVRT). CTA and MRA diagnostic results were obtained through qualitative assessments conducted by the Department of Radiology at Zhongda Hospital, whereas DVRT evaluations were conducted by two neurointerventional radiologists (each with >15 years of experience) from the Center of Interventional Radiology and Vascular Surgery. Two neurointerventional radiologists independently analyzed recurrence features (size, morphology, location) under blinded conditions, with Cohen’s kappa (κ) coefficient calculated as 0.968, indicating extremely high agreement. Initial discrepancies underwent adjudication by a third senior neurointerventional radiologist, thereby establishing a three-tier verification system. Our standardized 2D-DSA protocol, performed within 48 hours after routine CTA and MRA assessments, minimizes temporal variability in imaging comparisons. The 48-hour average interval in our study design not only prevents overestimation of diagnostic accuracy from delayed recurrence detection, but also aligns with established optimal post-embolization imaging windows (21,22). Imaging parameters for each modality were standardized as follows:

• CTA: performed on 256-detector CTA systems [Philips Brilliance iCT (Philips Healthcare, Amsterdam, the Netherlands) or United uCT960+ (United Imaging, Shanghai, China)] with settings: 120 kV/250 mA, slice thickness 0.5 mm, reconstruction interval 0.25 mm, and pitch 0.5. Multiplanar visualization was achieved through maximum intensity projection (MIP) and volume rendering technique (VRT).
• MRA: acquired using a 3.0 T MRI scanner (Philips Ingenia Elition X) with 3D time-of-flight (TOF) sequence, parameters: repetition time (TR) 21 ms/echo time (TE) 3.6 ms, slice thickness 0.5 mm, and matrix 320×320. Contrast enhancement was omitted to prevent interference from residual embolic materials.
• DVRT: data were acquired via Siemens Artis One 3D-DSA system (Siemens, Erlangen, Germany): 5-second acquisition with 20 mL iodixanol (320 mg I/mL) injected at 4 mL/s, capturing 133 projections. Dual-volume fusion (Syngo X Workplace) generated 0.2 mm³ isotropic voxel 3D models for morphological analysis.

Classification of recurrent aneurysms

In this study, recurrent aneurysms were systematically classified based on a comprehensive evaluation that incorporated data confirmed by 2D-DSA, a widely recognized gold standard for assessing ICA. The classification criteria included three primary dimensions: size, morphology, and location, each of which was meticulously analyzed to provide a detailed understanding of the recurrent aneurysms detected post-embolization (23,24):

• Size classification:
  • Small ICA, measuring less than 5 mm.
  • Medium ICA, ranging from 5 to 10 mm.
  • Large ICA, exceeding 10 mm in size.
• Morphological classification:
  • Saccular aneurysms, characterized by regular, smooth, sac-like structures.
  • Lobulated aneurysms, exhibiting multiple lobes, irregular margins, and a complex morphological profile.
  • Irregular aneurysms, distinguished by their irregular shape, blurred boundaries, and frequent presence of multiple filling defects.
  • Linear remnants, featuring a linear distribution, often adjacent to embolic materials or stents.
• Location classification:
  • Anterior circulation ICA, located on vessels such as the internal carotid artery, anterior cerebral artery, and middle cerebral artery.
  • Posterior circulation ICA, situated on the vertebral artery and basilar artery.

Evaluation of clinical outcomes

Patients’ clinical outcomes were meticulously documented during follow-up, encompassing aneurysm recurrence, necessity for re-embolization, cerebrovascular complications (including cerebral hemorrhage and ischemic stroke), and neurological function scores assessed using the modified Rankin Scale (mRS) (25,26). Recurrence was defined as a ≥1 mm increase in the size of the embolized aneurysm or the reappearance of contrast agents within the aneurysm neck or sac on subsequent imaging. When such recurrence was identified, re-embolization decisions were made by a multidisciplinary neurovascular team based on detailed imaging findings and comprehensive clinical risk stratification, with the aim of providing additional embolization procedures to patients deemed to require them.

Data processing and statistical analysis

Imaging data processing

All imaging data were meticulously examined through an independent, double-blind process to guarantee the absence of any resultant analytical overlap. Additionally, heatmaps and forest plots presented in Figures 1,2 were generated using Python (version 3.11.4; https://www.python.org/) utilizing the Matplotlib and Seaborn libraries for data visualization. These visualizations were based on the processed statistical outputs to enhance the clarity of comparative detection rates. Subsequent statistical analyses were rigorously conducted to assess the detection rates associated with CTA, MRA, and DVRT.

Figure 1 Comparative detection rates of vascular imaging modalities stratified by aneurysm characteristics. This heatmap illustrates the detection rates for three vascular imaging modalities across various aneurysm characteristics. The categories include aneurysm size (small, medium, large), morphology (saccular, lobulated, irregular), and location (anterior, posterior). The detection rates for each modality are expressed as percentages, with statistical significance indicated by asterisks (*, P<0.05; **, P<0.01) according to the P values provided. The heatmap color scale (from green to red) corresponds to detection rates, with green representing lower values and red representing higher detection rates. CTA, computed tomography angiography; DVRT, dual-volume reconstruction technology; MRA, magnetic resonance angiography.

Figure 2 Forest plot of detection rates with 95% CI for vascular imaging modalities in aneurysm characteristics. This forest plot compares the detection rates of three vascular imaging modalities across various aneurysm characteristics. The categories include aneurysm size (small, medium, large), morphology (saccular, lobulated, irregular, linear), and location (anterior, posterior). The detection rates are represented by markers: circles for DVRT, squares for CTA, and triangles for MRA, with horizontal error bars indicating the 95% CI for each detection rate. Each marker’s position reflects the point estimate of the detection rate, whereas the horizontal error bars show the range of the 95% CI, illustrating the variability and precision of the estimates. The width of the CIs visually indicates the reliability of the detection rate estimates. Narrower CI suggests higher precision, whereas wider intervals indicate greater uncertainty. This figure highlights the varying accuracy of each imaging modality across different aneurysm characteristics, with CIs providing insight into the precision and reliability of the detection rates. CI, confidence interval; CTA, computed tomography angiography; DVRT, dual-volume reconstruction technology; MRA, magnetic resonance angiography.

Analysis of the relationship between recurrent aneurysms and clinical outcomes

Multivariable logistic regression analysis was applied to examine the relationship between the size, morphology, and location of recurrent aneurysms and clinical outcomes. The impact of different recurrent aneurysm classifications on aneurysm recurrence and re-embolization was evaluated using logistic regression analysis.

Statistical methods

Data analysis was performed using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). Continuous variables were expressed as mean ± standard deviation, and comparisons between groups were conducted using t-tests or analysis of variance (ANOVA). Categorical variables were presented as frequencies and percentages, and comparisons between groups were performed using the Chi-squared test or Fisher’s exact test. Multivariable analysis was conducted using a logistic regression model. A P value <0.05 was considered statistically significant.

Results

Ultimately, a total of 152 patients who underwent ICA embolization between January 2020 and December 2023 met the inclusion criteria for this retrospective analysis. The cohort comprised 49 males (32.2%) and 103 females (67.8%), with a mean age of 58.9 years (range, 30–82 years). Below, we present the comparative detection performance of these imaging modalities, stratified by aneurysm size, morphology, and location, followed by an analysis of ICA recurrence risk factors.

Aneurysm detection and classification before embolization

Table 1 presents the detection rates of pre-embolization aneurysms, classified by size, morphology, and location, using three imaging modalities: CTA, MRA, and DVRT. The P values for the differences in detection rates between these modalities are also provided.

Aneurysm size classification

We compared detection performance across all size-based subgroups and performed inter-group comparisons of detection performance. For small aneurysms (<5 mm), DVRT demonstrated a detection rate of 88.7% (63/71), significantly outperforming CTA at 64.8% (46/71) and MRA at 59.2% (42/71) with a statistically significant P<0.001. For medium-sized aneurysms (5–10 mm), DVRT showed a detection rate of 92.6% (63/68), higher than those of CTA at 79.4% (54/68) and MRA at 73.5% (50/68) (P=0.001). In contrast, for large aneurysms (>10 mm), detection rates were high across all modalities, with DVRT at 92.3% (12/13), CTA at 84.6% (11/13), and MRA at 76.9% (10/13) (P=0.317).

Aneurysm morphological classification

For saccular aneurysms, DVRT had the highest detection rate of 95.3% (102/107), significantly outperforming CTA at 79.4% (85/107) and MRA at 72.0% (77/107) (P<0.001). For lobulated aneurysms, DVRT again showed superior detection at 91.9% (34/37) compared to CTA at 67.6% (25/37) and MRA at 64.9% (24/37) (P=0.003). For irregularly shaped aneurysms, detection rates were comparable across modalities: DVRT at 80.0% (4/5), CTA at 60.0% (3/5), and MRA at 40.0% (2/5) (P=0.400). For linear remnants, DVRT achieved perfect detection at 100.0% (3/3), whereas CTA and MRA detected only 33.3% (1/3) of cases, highlighting DVRT’s sensitivity for linear remnants (P=0.250).

Aneurysm location classification

For anterior circulation aneurysms, DVRT had a detection rate of 92.6% (100/108), significantly higher than those of CTA at 75.0% (81/108) and MRA at 69.4% (75/108) (P<0.001). For posterior circulation aneurysms, DVRT again showed the highest detection rate at 88.6% (39/44), compared to those of CTA at 68.2% (30/44) and MRA at 63.6% (28/44) (P=0.005). DVRT outperformed both CTA and MRA in detecting aneurysms across various sizes, morphologies, and locations. It demonstrated particularly high sensitivity for small (<5 mm), saccular, and lobulated aneurysms, as well as for aneurysms located in both anterior and posterior circulations. These findings underscore DVRT’s potential as a sensitive imaging modality for pre-embolization aneurysm detection, which may aid clinical decision-making and improve patient outcomes.

Comparison of detection accuracy for recurrent aneurysms

Pre-embolization aneurysm characteristics are summarized in Table 1. Table 2 compares the detection accuracy for recurrent aneurysms across three imaging modalities—CTA, MRA, and DVRT—categorized by aneurysm size, morphology, and location. In addition, Table 3 presents the results of a logistic regression analysis identifying pre-embolization characteristics associated with aneurysm recurrence risk. This analysis provides a comprehensive breakdown of the factors influencing recurrence, including aneurysm size, morphology, and location, which are critical in guiding clinical decision-making for postoperative surveillance and management.

Overall recurrent aneurysm detection

DVRT demonstrated a significantly higher a detection rate of 94.4% (17/18), significantly higher than those of CTA (77.8%, 14/18) and MRA (50.0%, 9/18) (P<0.001), confirming its superior sensitivity for recurrent ICA.

Aneurysm size classification

For small aneurysms (<5 mm), DVRT achieved 80.0% (4/5), outperforming CTA (60.0%, 3/5) and MRA (40.0%, 2/5) (P=0.317). In medium aneurysms (5–10 mm), DVRT reached 100.0% (9/9), significantly outperforming CTA (88.9%, 8/9) and MRA (55.6%, 5/9) (P=0.012). For large aneurysms (>10 mm), DVRT achieved 100.0% (4/4) detection rate for large aneurysms (>10 mm), significantly higher than those of CTA (75.0%, 3/4) and MRA (50.0%, 2/4), although the difference was not statistically significant (P=0.250). The heatmap (Figure 1) illustrates these differences, highlighting DVRT’s superior detection rates across all aneurysm sizes. DVRT maintained high detection consistency, whereas CTA and MRA showed reduced sensitivity, particularly for small aneurysms.

Aneurysm morphological classification

For saccular aneurysms, DVRT achieved 100.0% (9/9) detection, significantly higher than that of CTA (88.9%, 8/9) and MRA (66.7%, 6/9) (P=0.039). For lobulated aneurysms, DVRT also showed perfect detection (100.0%, 4/4), outperforming CTA (75.0%, 3/4) and MRA (50.0%, 2/4), though the difference was not statistically significant (P=0.250). For irregular aneurysms, detection rates were similar between DVRT (66.7%, 2/3) and CTA (66.7%, 2/3), whereas MRA had lower sensitivity (33.3%, 1/3) (P>0.99). For linear remnants, DVRT maintained perfect detection (100.0%, 2/2), significantly higher than that of CTA (50.0%, 1/2) and MRA (0.0%, 0/2), though the difference was not significant (P=0.400). The forest plot displayed in Figure 2 reinforces these findings, showing narrow confidence intervals (CIs) for DVRT, particularly in saccular and lobulated aneurysms. CTA and MRA displayed broader variability, especially in irregular and linear aneurysms.

Aneurysm location classification

For anterior circulation aneurysms, DVRT achieved a detection rate of 92.3% (12/13), significantly outperforming CTA (76.9%, 10/13) and MRA (46.2%, 6/13) (P=0.008). In posterior circulation aneurysms, DVRT demonstrated a better detection rate of 100.0% (5/5), compared to those of CTA (80.0%, 4/5) and MRA (60.0% 3/5), although this difference did not reach statistical significance (P=0.200). The heatmap in Figure 1 clearly illustrates DVRT’s superior detection rates (shown in dark red) across both anterior and posterior circulation aneurysms compared to CTA (orange) and MRA (light green/yellow). The color gradient effectively demonstrates DVRT’s consistent high performance, particularly for posterior circulation aneurysms where it achieved 100% (5/5) detection versus CTA’s 80% (4/5) and MRA’s 60% (3/5).

Relationship between recurrent aneurysms and clinical outcomes

Having established DVRT’s superior detection performance, we further analyzed the clinical implications of aneurysm recurrence.

Recurrent aneurysm size and clinical outcomes

Recurrence rates significantly differed by aneurysm size (P=0.046), with small aneurysms (<5 mm) demonstrating a recurrence rate of 7.0% (5/71), medium aneurysms (5–10 mm) 13.2% (9/68), and large aneurysms (>10 mm) 30.8% (4/13). Multivariable logistic regression identified aneurysm size as an independent predictor of recurrence: large aneurysms (>10 mm) exhibited a 2.5-fold increased risk compared to small aneurysms [odds ratio (OR) =2.51, 95% CI: 1.83–3.51, P=0.030], whereas medium aneurysms (5–10 mm) showed a 1.8-fold higher risk (OR =1.83, 95% CI: 1.22–2.73, P=0.011). These findings underscore the critical role of aneurysm size in recurrence risk stratification, with larger aneurysms requiring closer postoperative surveillance.

Recurrent aneurysm morphology and clinical outcomes

Recurrence rates exhibited significant variation across morphological subtypes (P<0.001). Saccular aneurysms demonstrated the lowest recurrence rate at 8.4% (9/107), followed by lobulated aneurysms at 10.8% (4/37). In contrast, irregularly shaped aneurysms showed substantially higher recurrence rates of 60.0% (3/5), whereas linear remnants exhibited the highest recurrence rate at 66.7% (2/3). Multivariable analysis revealed that irregular aneurysms had a significantly higher recurrence risk (OR =3.02, 95% CI: 2.15–4.43, P=0.003). Similarly, lobulated aneurysms also showed an increased re-embolization risk (OR =2.25, 95% CI: 1.54–3.21, P=0.005). These findings highlight the critical influence of morphological complexity on treatment outcomes, with irregular and lobulated subtypes requiring particularly vigilant postoperative monitoring.

Recurrent aneurysm location and clinical outcomes

The location of the aneurysm also plays a critical role in recurrence risk. In our study, the recurrence rates of anterior (12.0%, 13/108) and posterior circulation aneurysms (11.4%, 5/44) showed no significant difference (P>0.99). However, Fisher’s exact test revealed that DVRT had significantly higher detection rates for anterior circulation aneurysms compared to CTA and MRA (P<0.001). Logistic regression analysis, after controlling for potential confounding factors, indicated that posterior circulation aneurysms had a significantly higher recurrence risk, with an OR of 2.07 (95% CI: 1.43–2.91, P<0.001). This suggests that, although the recurrence rates were similar between the two groups, posterior circulation aneurysms are at a higher risk of recurrence. This may be attributed to specific characteristics of posterior circulation aneurysms, such as their anatomical location, hemodynamics, and treatment challenges, which make them more prone to recurrence. Therefore, patients with posterior circulation aneurysms require closer monitoring and more aggressive follow-up strategies.

Multivariable logistic analysis

Our multivariable logistic regression analysis identified three key independent risk factors for aneurysm recurrence: aneurysm size, morphology, and location. Larger aneurysms were associated with an increased risk of recurrence, with an OR of 2.51 (P=0.030), highlighting the importance of achieving complete aneurysm occlusion during treatment. Specifically, larger aneurysms were more prone to recurrence compared with smaller aneurysms. This is further supported by the recurrence rate findings in Table 2, where large aneurysms (>10 mm) demonstrated a recurrence rate of 30.8% (4/13), significantly higher than that for small aneurysms (7.0%, 5/71) (P=0.046).

Aneurysm morphology emerged as another significant predictor of recurrence. Irregularly shaped aneurysms exhibited an OR of 3.02 (P=0.003), indicating that these complex aneurysms are at a higher risk of recurrence. This finding is consistent with the higher recurrence rate for irregularly shaped aneurysms, which reached 60.0% (3/5), compared to 8.4% (9/107) for saccular aneurysms (Table 2). The challenges associated with irregular morphology further underscore the necessity for more aggressive treatment strategies or closer follow-up in these patients. Location also was also shown to play a critical role in recurrence risk, with posterior circulation aneurysms demonstrating a significantly increased risk of recurrence (OR =2.07, P<0.001). As shown in Table 2, posterior circulation aneurysms had a recurrence rate of 11.4% (5/44), compared to 12.0% (13/108) for anterior circulation aneurysms, suggesting that despite a similar recurrence rate, posterior circulation aneurysms require more tailored management strategies due to their inherent challenges. These findings are detailed in Tables 2,3, which offer a comprehensive overview of the recurrence risks associated with aneurysm characteristics. The multivariable logistic analysis provides crucial insights into the factors that influence aneurysm recurrence and emphasizes the importance of personalized management approaches.

Discussion

The comprehensive evaluation of recurrent ICA necessitates a multimodal imaging approach, particularly in the context of post-embolization surveillance. By leveraging high-resolution, multi-angle reconstruction, DVRT not only enhances the visualization of ICA morphology but also provides critical insights into the dynamic interplay between anatomical characteristics and recurrence risk. The following section will first delve into the detection efficacy of three different imaging techniques (CTA, MRA, and DVRT) for the characteristics of aneurysms, and then explore which characteristics of preoperative ICA will have impact on postoperative recurrence and clinical outcomes.

Association between aneurysm size and recurrence risk

Since aneurysms tend to increase in size over time, and larger size is associated with higher risk of rupture, follow-up imaging to assess the growth of aneurysms is recommended for aneurysms that are left untreated (27,28). Our study further extends this rationale to treated aneurysms by demonstrating a clear relationship between aneurysm size and recurrence risk post-embolization. Small aneurysms (<5 mm) exhibited the lowest recurrence rate (7.0%, 5/71), whereas large aneurysms (>10 mm) showed a significantly higher recurrence rate (30.8%, 4/13) (P=0.046). Figure 1 highlights post-embolization DVRT’s detection superiority for small aneurysms (80%, 60%, and 40% for DVRT, CTA, and MRA, respectively), which are prone to being overlooked by conventional modalities (CTA, MRA). Figure 2 demonstrates DVRT’s superior sensitivity with the narrowest CIs in detecting recurrent ICA across all size categories. Notably, in medium-sized aneurysms, a partial overlap was observed between the DVRT and CTA CIs (95% CI: 82–98% vs. 75–93%, respectively), suggesting potential statistical equivalence for aneurysms with specific morphological characteristics. Furthermore, the overlapping CIs between CTA and MRA for both large (CTA: 68–92%; MRA: 55–85%) and small aneurysms (CTA: 53–77%; MRA: 38–62%) imply that distinct morphological characteristics (e.g., shape complexity) and anatomical locations of the ICA may modulate detection performance, potentially equalizing the diagnostic capabilities of different imaging modalities within particular subgroups. Logistic regression analysis confirmed that larger aneurysms (>10 mm) had 2.51 times higher odds of recurrence compared to small aneurysms (95% CI: 1.83–3.51, P=0.030). This aligns with previous studies highlighting the challenges of achieving complete occlusion in larger aneurysms due to complex hemodynamics and residual perfusion risks (29-31). For instance, Fujii et al. (29) reported similar trends in their cohort of large internal carotid artery aneurysms treated with flow diverters, where incomplete embolization was a key predictor of recurrence. Larger aneurysms often exhibit turbulent flow patterns and wider necks, which impede stable coil placement and increase the likelihood of recanalization (30). Additionally, Wang et al. (12) emphasized that incomplete endothelialization over larger aneurysm surfaces may prolong exposure to hemodynamic stress, further elevating recurrence risks. These findings underscore the importance of adjunctive techniques, such as stent-assisted coiling or flow diversion, in managing large aneurysms to mitigate recurrence risks (31,32).

Association between aneurysm morphology and recurrence risk

ICA morphology significantly influenced recurrence outcomes. Figure 1 demonstrates DVRT’s superior detection of lobulated (100.0% detection rate, 4/4) and irregular aneurysms (66.7%, 2/3), underscoring its clinical value in identifying high-risk morphologies that may require re-intervention. Figure 2 confirms DVRT’s superior sensitivity with the narrowest CIs across all morphological categories. For saccular aneurysms, partial overlap between DVRT and CTA CIs (95% CI: 95–100% vs. 85–95%) suggests comparable detection rates for simpler morphologies. The overlapping CIs between CTA and MRA for both lobulated (CTA: 70–85%; MRA: 50–75%) and irregular aneurysms (CTA: 55–78%; MRA: 25–42%) indicate that morphological complexity may modulate detection performance.

Furthermore, irregularly shaped aneurysms exhibited a 60.0% (3/5) recurrence rate (P<0.001 vs. saccular aneurysms), with logistic regression showing a 3.02-fold increased risk (95% CI: 2.15–4.43, P=0.003). These results are consistent with Liu et al. (33), who attributed high recurrence rates in irregular aneurysms to uneven coil distribution and residual neck remnants. The geometric complexity of irregular aneurysms often creates “blind zones” where blood flow persists despite embolization, promoting coil compaction and recurrence (34). Marbacher et al. (35) further demonstrated that lobulated aneurysms are prone to residual perfusion in secondary lobes, which may evade detection by conventional imaging. Notably, Hashimoto et al. (36) highlighted that irregular aneurysms often correlate with intraluminal thrombus formation, which can shield residual flow from conventional angiographic visualization. DVRT’s multi-angle reconstruction capabilities may overcome this limitation by enhancing boundary delineation (21).

Association between aneurysm location and recurrence risk

Posterior circulation aneurysms exhibited a 2.07-fold higher recurrence risk than anterior circulation lesions (95% CI: 1.43–2.91, P<0.001), attributable to their complex hemodynamic environment. As reported by Hu et al. (37), the vertebrobasilar system’s tortuous anatomy and bidirectional flow patterns create low shear stress that promotes endothelial proliferation, whereas Yuan et al. (17) identified that pulsatile basilar artery flow generates mechanical “water-hammer effects”, which may allow blood flow to enter the embolized area, thereby increasing recurrence risk. Conventional imaging modalities showed limited detection efficacy in this region (CTA: 80.0%; MRA: 60.0%) (Table 2 and Figure 1). This limitation stems from the V4 segment’s proximity to dense osseous structures (e.g., clivus, occipital condyles, and foramen magnum), which predisposes it to significant bone-induced artifacts during vascular imaging due to inherent technical constraints—bone artifacts in CTA and flow-related signal loss in MRA (13,14). In contrast to posterior circulation findings, anterior circulation recurrent aneurysms showed significant variability in detection rates across imaging modalities (Table 2). DVRT demonstrated the highest sensitivity (92.3%, 12/13), followed by CTA (76.9%, 10/13) and MRA (46.2%, 6/13; P=0.008). This observed hierarchy of diagnostic performance should be interpreted with caution, as van Niftrik et al.’s study of anterior circulation moyamoya disease revealed only moderate agreement between MRA and DSA, despite the theoretical completeness of magnetic resonance (MR)-based assessment (38). Notably, Dong et al. (39) emphasized that in the context of stent-assisted coiling for unruptured anterior circulation aneurysms, conventional imaging modalities including DSA, CTA, and MRA remain the cornerstone for mid-to-long-term follow-up. Particularly in cases complicated by in-stent stenosis (ISS), their findings reinforce DSA’s continued primacy as the preferred surveillance method, owing to its superior spatial resolution in detecting vascular wall irregularities within stent architecture.

In contrast, DVRT achieved 100.0% (5/5) detection for posterior circulation aneurysms in our cohort (Table 2, Figures 1,2), demonstrating particular value for high-risk locations (such as posterior circulation aneurysms) where traditional surveillance often underestimates recurrence. Li et al. (21) demonstrated that DVRT’s dual-volume processing minimizes metallic artifacts from embolization coils, whereas Zheng et al. (11) validated its spatial resolution in differentiating coils from residual aneurysms, a finding corroborated by our heatmap analysis (Figures 1,2). DVRT’s multi-planar reconstruction capability enables precise visualization of coil-sac spatial relationships, facilitating identification of critical recurrence patterns such as “dog-ear” neck remnants (18). This is exemplified in Figure 3 (Patient A), where DVRT clearly delineates the spatial relationship between coils and residual aneurysm sac. These characteristics support the clinical utility of DVRT compared with CTA and MRA for monitoring posterior circulation aneurysms, particularly in clinical scenarios requiring serial imaging to manage recurrence risks (40,41).

Figure 3 Images of 4 patients with intracranial aneurysms evaluated using 2D-DSA, 3D-DSA, and DVRT. Patients A–D, each with aneurysms located in different regions, underwent endovascular embolization treatment. Post-treatment outcomes are favorable in the 2D-DSA images. However, the 3D-DSA images reveal various morphological features of aneurysm recurrence or residual sacs. Specifically, patient A’s aneurysm is located in the right vertebral artery V4 segment, patient B’s aneurysm is in the right internal carotid artery C5 segment, patient C’s aneurysm is at the origin of the left posterior communicating artery, and patient D’s aneurysm is in the right internal carotid artery C1 segment. The DVRT images utilize distinct color coding to differentiate recurrent aneurysms or residual sacs from coils and the parent artery, offering clear visualization of their spatial relationships. White arrows mark the embolized aneurysm at 1-year follow-up. Purple regions in DVRT denote occluded (non-flowing) portions post-coiling, while other colors indicate residual flow. 2D-DSA, two-dimensional digital subtraction angiography; 3D-DSA, three-dimensional digital subtraction angiography; DVRT, dual-volume reconstruction technology.

Clinical implications and DVRT

From a clinical perspective, our study findings highlight the advantages of DVRT in detecting the multifactorial nature of ICA recurrence. The multivariable logistic regression explicitly incorporated pre-embolization aneurysm characteristics (size, morphology, location) to adjust for their predictive value for recurrence, independent of post-procedural changes. Our findings also underscore the importance of a personalized follow-up strategy that takes into account the specific characteristics of each patient’s aneurysm, including its size, morphology, and location, as well as utilizing advanced imaging modalities such as DVRT to monitor for recurrence. In clinical practice, this comprehensive understanding of risk factors can guide clinicians in determining the frequency of follow-up imaging, the selection of appropriate imaging modalities, and the design of tailored treatment plans for patients at high risk of recurrence.

Limitations

This study has limitations: its retrospective, single-center design may restrict generalizability, as institutional practices (e.g., coil types, embolization techniques) could influence recurrence rates. Although the protocol specified standardized follow-up intervals, actual follow-up duration varied among patients, with a small subset deviating to 5 or 16 months (vs. the scheduled 3- and 12-month intervals) due to logistical constraints, which may introduce survivor bias and affect recurrence rate calculations. A statistical power analysis for recurrent aneurysm samples showed that power approached but did not reach 80%, largely due to small sample sizes of linear remnants, which may have hindered accurate reflection of detection rates across the three imaging modalities. Although DVRT demonstrated superior accuracy, its cost-effectiveness compared to that of CTA/MRA remains unaddressed—a critical gap for resource-limited settings. Future multicenter prospective studies with standardized protocols are essential to validate DVRT’s long-term diagnostic accuracy and establish a cost-benefit framework.

Conclusions

DVRT demonstrates superior diagnostic accuracy over CTA and MRA in detecting recurrent ICAs post-embolization, particularly for small (<5 mm), irregularly shaped, and posterior circulation aneurysms. Our analysis identifies aneurysm size (>10 mm), irregular morphology, and posterior circulation location as independent risk factors for recurrence. DVRT’s high-resolution, multi-angle imaging enables precise morphological delineation and early detection of subtle recurrences, supporting risk stratification and personalized postoperative surveillance. Integrating DVRT into routine follow-up protocols may optimize clinical decision-making, facilitate timely interventions, and improve long-term outcomes for high-risk patients.

Acknowledgments

The data collection work for this study benefited from the active participation and assistance of Supervisor Nurse Yuehong Ding and Hongyan Wan. By utilizing their professional knowledge and practical experience, they ensured the accuracy and completeness of the data, thereby laying a solid foundation for the subsequent analysis of the study.

Footnote

Reporting Checklist: The authors completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-303/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-303/dss

Funding: This work was supported by the National Key Technologies R&D Program (specifically under grant number 2024YFB4708804), and the Jiangsu Province Science and Technology Support Project (grant number BE2023769).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-303/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 Ethics Committee of Zhongda Hospital Affiliated to Southeast University (No. 2020ZDSYLL158-P01) and the requirement for 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. Juvela S. Outcome of Patients with Multiple Intracranial Aneurysms after Subarachnoid Hemorrhage and Future Risk of Rupture of Unruptured Aneurysm. J Clin Med 2021;10:1712.
2. Rinkel GJ, Ruigrok YM. Preventive screening for intracranial aneurysms. Int J Stroke 2022;17:30-6.
3. Jou LD, Desai VR, Britz GW. In vitro investigation of contrast flow jet timing in patient-specific intracranial aneurysms. Quant Imaging Med Surg 2016;6:134-43.
4. Molenberg R, Aalbers MW, Appelman APA, Uyttenboogaart M, van Dijk JMC. Intracranial aneurysm wall enhancement as an indicator of instability: a systematic review and meta-analysis. Eur J Neurol 2021;28:3837-48.
5. Fu M, Peng F, Zhang M, Chen S, Niu H, He X, Xu B, Liu A, Li R. Aneurysmal wall enhancement and hemodynamics: pixel-level correlation between spatial distribution. Quant Imaging Med Surg 2022;12:3692-704.
6. Lu Y, Wang C, Bao Y, Huang L, Lu G, Li Y. Association between intracranial aneurysm wall enhancement and intracranial atherosclerotic plaque: a cross-sectional study using high-resolution vessel wall imaging. Quant Imaging Med Surg 2024;14:1553-63.
7. Wang Y, Gao X, Lu A, Zhou Z, Li B, Sun X, Zhu B. Residual aneurysm after metal coils treatment detected by spectral CT. Quant Imaging Med Surg 2012;2:137-8.
8. Egashira Y, Yoshimura S, Enomoto Y, Ishiguro M, Asano T, Iwama T. Ultra-early endovascular embolization of ruptured cerebral aneurysm and the increased risk of hematoma growth unrelated to aneurysmal rebleeding. J Neurosurg 2013;118:1003-8.
9. Wang H, Li W, He H, Luo L, Chen C, Guo Y. 320-detector row CT angiography for detection and evaluation of intracranial aneurysms: comparison with conventional digital subtraction angiography. Clin Radiol 2013;68:e15-20.
10. Li Q, Lv F, Yao G, Li Y, Xie P. 64-section multidetector CT angiography for evaluation of intracranial aneurysms: comparison with 3D rotational angiography. Acta Radiol 2014;55:840-6.
11. Zheng H, Zhang X, Gui S, Yang M, Wang J. Performance of S-MAR in metal artifact reduction for intracranial aneurysm patients after endovascular embolization: a cross-sectional spectral computed tomography study. Quant Imaging Med Surg 2025;15:4235-46.
12. Wang Y, Zhao L, Zhang X, Zheng J, Geng Y, Huang B, Chen T, Qiang J, Liu B, Zhang L, Zhang X. Intracranial aneurysm rupture risk in northern China: a retrospective case-control study. Quant Imaging Med Surg 2024;14:376-85.
13. Li MH, Li YD, Gu BX, Cheng YS, Wang W, Tan HQ, Chen YC. Accurate diagnosis of small cerebral aneurysms ≤5 mm in diameter with 3.0-T MR angiography. Radiology 2014;271:553-60.
14. Thaker NG, Turner JD, Cobb WS, Hussain I, Janjua N, He W, Gandhi CD, Prestigiacomo CJ. Computed tomographic angiography versus digital subtraction angiography for the postoperative detection of residual aneurysms: a single-institution series and meta-analysis. J Neurointerv Surg 2012;4:219-25.
15. Adeeb N, Griessenauer CJ, Patel AS, Moore J, Dolati-Ardejani P, Gupta R, Motiei-Langroudi R, Ogilvy CS, Thomas AJ. Reliability of dual- vs single-volume reconstruction of three-dimensional digital subtraction angiography for follow-up evaluation of endovascularly treated intracranial aneurysms. Interv Neuroradiol 2016;22:687-92.
16. Lee HJ, Choi JH, Kim BS, Shin YS. Dual Volume Image Fusion Technique for Anatomical Evaluation of Cavernous Sinus Dural Arteriovenous Fistula: Usefulness for Transvenous Target Segment Embolization. World Neurosurg 2024;190:e181-90.
17. Yuan J, Huang C, Li Z, Jiang X, Zhao X, Lai N, Xia D, Wu D, Zhang B, Wang X, Fang X. Hemodynamic Characteristics Associated with Recurrence of Middle Cerebral Artery Bifurcation Aneurysms After Total Embolization. Clin Interv Aging 2021;16:2023-32.
18. Fuga M, Tanaka T, Irie K, Kajiwara I, Tachi R, Teshigawara A, Ishibashi T, Hasegawa Y, Murayama Y. Proposed predictors of the need for retreatment after coil embolization of unruptured cerebral aneurysms with major or minor recanalization: Analysis of a single center’s experience. Interdisciplinary Neurosurgery 2021;26:101296.
19. Society of Neurosurgery of Chinese Medical Association. National Center for Neurological Disorders; National Clinical Research Center for Neurological Diseases. Chinese guideline for the clinical management of patients with unruptured intracranial aneurysms (2024). Zhonghua Yi Xue Za Zhi 2024;104:1918-39.
20. Society of Neurosurgery of Chinese Medical Association. National Center for Neurological Disorders; National Clinical Research Center for Neurological Diseases. Chinese guideline for the clinical management of patients with ruptured intracranial aneurysms (2024). Zhonghua Yi Xue Za Zhi 2024;104:1940-71.
21. Li B, Zhuo J, Song G, Zhang J, Zhao W, Zhao Z, Wang L, Dong Y, Zhang X, Zhang G. Single-Center Analysis of Reliability of 3D-DSA Dual Volume Reconstruction for Evaluation of Endovascularly Treated Intracranial Aneurysms. World Neurosurg 2021;155:e150-5.
22. Greffier J, Dabli D, Hamard A, Akessoul P, Belaouni A, Beregi JP, Frandon J. Impact of dose reduction and the use of an advanced model-based iterative reconstruction algorithm on spectral performance of a dual-source CT system: A task-based image quality assessment. Diagn Interv Imaging 2021;102:405-12.
23. Yang J, Wu Q, Yuan K, Zhang G, Xu C, Dai J, Du Y, Sun B, Zhu Y, Lan S, Shi H, Xu S. Reconstructive endovascular treatment of patients with ruptured distal intracranial aneurysms in posterior circulation: a retrospective cohort study. Quant Imaging Med Surg 2024;14:8878-88.
24. Lu C, Pan XP, Zhao ZH, Zhang Y. Interventional Radiology Basics (3rd ed.). Beijing: People's Medical Publishing House; 2020.
25. Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J, Holman RInternational Subarachnoid Aneurysm Trial (ISAT) Collaborative Group. International Subarachnoid Aneurysm Trial (ISAT) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: a randomised trial. Lancet 2002;360:1267-74.
26. Spetzler RF, McDougall CG, Zabramski JM, Albuquerque FC, Hills NK, Russin JJ, Partovi S, Nakaji P, Wallace RC. The Barrow Ruptured Aneurysm Trial: 6-year results. J Neurosurg 2015;123:609-17.
27. Jin D, Song C, Leng X, Han P. A systematic review and meta-analysis of risk factors for unruptured intracranial aneurysm growth. Int J Surg 2019;69:68-76.
28. van Lieshout JH, Mijderwijk HJ, Nieboer D, Lingsma HF, Ahmadi SA, Karadag C, Muhammad S, Porčnik A, Wasilewski D, Wessels L, van Donkelaar CE, van Dijk JMC, Hänggi D, Boogaarts HD. Development and Internal Validation of the ARISE Prediction Models for Rebleeding After Aneurysmal Subarachnoid Hemorrhage. Neurosurgery 2022;91:450-8.
29. Fujii T, Teranishi K, Yatomi K, Suzuki K, Mitome-Mishima Y, Kondo A, Oishi H. Long-term Follow-up Results after Flow Diverter Therapy Using the Pipeline Embolization Device for Large or Giant Unruptured Internal Carotid Artery Aneurysms: Single-center Retrospective Analysis in the Japanese Population. Neurol Med Chir (Tokyo) 2022;62:19-27.
30. Lee W, Han HJ, Kim J, Park KY, Kim YB, Jang CK, Chung J. Flow diverter for the treatment of large (> 10 mm) vertebral artery dissecting aneurysms. Acta Neurochir (Wien) 2022;164:1247-54.
31. Jin J, Guo G, Ren Y, Yang B, Wu Y, Wang S, Sun Y, Wang X, Wang Y, Zheng J. Risk Factors for Recurrence of Intracranial Aneurysm After Coil Embolization: A Meta-Analysis. Front Neurol 2022;13:869880.
32. Guo H, Liu JF, Li CH, Wang JW, Li H, Gao BL. Effects of stent-assisted coiling in comparison with flow diversion on intracranial aneurysms. Front Neurol 2022;13:937536.
33. Liu Q, Yang Y, Yang J, Li M, Yang S, Wang N, Wu J, Jiang P, Wang S. Rebleeding of Ruptured Intracranial Aneurysm After Admission: A Multidimensional Nomogram Model to Risk Assessment. Front Aging Neurosci 2021;13:692615.
34. Uchikawa H, Kin T, Koizumi S, Sato K, Uchida T, Takeda Y, Koike T, Kiyofuji S, Yamashiro S, Mukasa A, Saito N. Aneurysmal Inflow Rate Coefficient Predicts Ultra-early Rebleeding in Ruptured Intracranial Aneurysms: Preliminary Report of a Computational Fluid Dynamics Study. Neurol Med Chir (Tokyo) 2023;63:450-6.
35. Marbacher S, Grüter BE, Wanderer S, Andereggen L, Cattaneo M, Trost P, Gruber P, Diepers M, Remonda L, Steiger HJ. Risk of intracranial aneurysm recurrence after microsurgical clipping based on 3D digital subtraction angiography. J Neurosurg 2023;138:717-23.
36. Hashimoto Y, Matsushige T, Kawano R, Hara T, Kobayashi S, Kaneko M, Ono C, Horie N. High signal intensity of the intraaneurysmal sac on T1 CUBE imaging as a predictor of aneurysm stability after coil embolization. J Neurosurg 2024;140:144-52.
37. Hu X, Deng P, Ma M, Tang X, Qian J, Wu G, Gong Y, Gao L, Zou R, Leng X, Xiang J, Wu J, Ding Z. How does the recurrence-related morphology characteristics of the Pcom aneurysms correlated with hemodynamics? Front Neurol 2023;14:1236757.
38. van Niftrik CHB, Sebök M, Nicholson P, Olijnyk L, Thurner P, Venkatraghavan L, Schaafsma J, Radovanovic I, Fisher JA, Krings T, Kulcsár Z, Tymianski M, Regli L, Mikulis DJ, Fierstra J. A dual-center validation of the PIRAMD scoring system for assessing the severity of ischemic Moyamoya disease. Quant Imaging Med Surg 2023;13:4618-32.
39. Dong L, Wang C, Wu X, Xu H, Wu X, Zhang Y, Lv M. Midterm outcomes and costs of Pipeline embolization device alone versus Atlas stent-assisted coiling for unruptured anterior circulation aneurysms: a propensity score matched comparative analysis. Quant Imaging Med Surg 2025;15:1977-89.
40. Vaphiades MS, Cure J, Kline L. Management of intracranial aneurysm causing a third cranial nerve palsy: MRA, CTA or DSA? Semin Ophthalmol 2008;23:143-50.
41. Xu Q, Tian Q, Han W, Liu C, Liao J, Li M. Endovascular coiling versus microsurgical clipping for extremely small intracranial aneurysms: a comparative analysis of treatment strategies, complications, and clinical outcomes. Quant Imaging Med Surg 2025;15:5604-20.