Quantitative digital subtraction angiography for identifying the risk factors of incomplete occlusion in intracranial aneurysms with branch incorporation after flow diverter treatment

Introduction

Flow diverters (FDs) have revolutionized the treatment of intracranial aneurysms (IAs), achieving complete occlusion rates of 75–85%, with most cases reaching a plateau at 12 to 18 months (1,2). However, persistent incomplete occlusion remains a critical concern, particularly for aneurysms involving branch arteries, with branch incorporation observed in nearly 40–50% of cases with incomplete occlusion (1,3). Branch incorporation has been recognized in several studies to be a significant factor influencing incomplete occlusion, yet the underlying mechanisms remain inadequately clarified (1,4). Some studies have suggested that pressure gradients maintain blood flow through incorporated branches, thereby preventing aneurysm occlusion (5). However, evidence from quantitative hemodynamics remains lacking. Therefore, investigating aneurysms involving branches requires not only identifying the factors that contribute to incomplete occlusion in this specific subset but also clarifying the role of branch blood flow in the underlying mechanisms of occlusion.

Recent research in this area has focused on hemodynamics, as FD treatment is highly related to flow reconstruction, and thus the study of hemodynamic factors in incomplete occlusion is crucial (6,7). Quantitative digital subtraction angiography (QDSA) has become a valuable technique that can estimate hemodynamics through angiographic parametric imaging (API) variables (8,9). Early studies by Doerfler et al. demonstrated the feasibility of using QDSA techniques to assess the hemodynamic changes induced by FDs in both animal models and patients, supporting its potential to predict aneurysm occlusion (10,11). Liang et al. further confirmed its value in predicting occlusion outcomes in FD-treated IAs (12). Additionally, QDSA analysis has been used to demonstrate the positive impact of FDs on blood flow in covered branches (9). However, there is limited research on the hemodynamic factors associated with incomplete occlusion specifically in aneurysms with branch incorporation. Moreover, FD treatment is often combined with coiling in complex aneurysms, such as those involving branch arteries, but whether coiling affects branch hemodynamics and contributes to better occlusion remains unclear.

Therefore, we aimed to identify the risk factors of incomplete occlusion in IAs with branch incorporation using QDSA analysis and used subgroup analysis to further examine whether coiling has an impact on branch hemodynamics and contributes to improved occlusion outcomes. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-358/rc).

Methods

Study design and patient selection

This retrospective study was approved by the Institutional Review Board of Beijing Tiantan Hospital (approval ID: KY2023-261-01) and was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Due to the retrospective nature of the study, the requirement for written informed consent from participants was waived. We analyzed data from 111 patients with 113 IAs involving incorporated branches who underwent FD treatment between January 2016 and December 2022. The inclusion criteria were as follows: (I) age 18–80 years; (II) FD treatment; (III) at least one digital subtraction angiography (DSA) follow-up visit after treatment; and (IV) a branch artery originating from the aneurysm dome or neck. The following exclusion criteria were applied: (I) a DSA follow-up duration shorter than 1 year; (II) coexisting cerebrovascular disorders including arteriovenous malformations, fistula, or moyamoya disease; (III) suboptimal image quality that compromised morphological assessments or QDSA evaluations (e.g., motion artifacts, insufficient contrast filling, or inability to expose the entire IA); and (IV) IAs without branch involvement.

Based on the occlusion outcomes evaluated via follow-up DSA, the 113 branch-incorporated IAs were finally divided into an incomplete occlusion group (n=52) and a complete occlusion group (n=61). For evaluating aneurysm occlusion outcomes, complete occlusion was defined as an O’Kelly-Marotta grade D, while and incomplete occlusion was considered to be O’Kelly-Marotta grades A to C (2,13). All angiographic images were independently reviewed and analyzed by an expert panel of neurosurgeons.

Clinical and aneurysm-related characteristics

Data were collected on patient demographics (age and sex), medical history (hypertension, coronary heart disease, diabetes mellitus, smoking, and drinking), and aneurysm-related features (ruptured status, aneurysm site, size, neck diameter, parent artery diameter, saccular shape, and treatment type). Measurements also included branch characteristics, specifically, branch diameter and whether it originated from the dome or neck.

QDSA analysis and hemodynamic variables

Image preparation and time-density curve (TDC) construction

The schematic workflow of QDSA analysis is shown in Figure 1. DSA images were acquired from Philips Healthcare, Siemens Healthineers, and GE HealthCare equipment in Digital Imaging and Communications in Medicine format and processed in MATLAB software (MathWorks, Natick, MA, USA). For each DSA acquisition, Visipaque (GE HealthCare, Chicago, IL, USA) was power-injected via intermediate catheters into the internal carotid artery (flow rate: 4 mL/s, volume: 6 mL) or vertebral artery (flow rate 3 mL/s, volume 5 mL) (12). The immediate postoperative images included the entire hemodynamic phase from initial contrast arrival through the venous sinus phase, facilitating the capture of critical flow dynamics. The most suitable projection was selected for each aneurysm to optimize visualization and reduce overlapping vessels (14). TDCs were derived pixel by pixel from the DSA sequences to reflect changes in contrast concentration over time. To correct for background interference and standardize across participants, the values of grayscale intensity were normalized by initial level. Raw TDCs were denoised via fast Fourier transform, which filtered low-power frequency noise while retaining signal patterns of physiological relevance (15). The denoised TDCs were subsequently reconstructed back into the time domain and aligned temporally with the contrast arrival time being set to zero, thereby establishing a consistent reference point for further analysis.

Figure 1 Schematic workflow of quantitative digital subtraction angiography analysis. (A) Image preprocessing: consecutive two-dimensional digital subtraction angiography sequences were aligned on the optimal working projection to simultaneously visualize the aneurysm sac and incorporated branch. (B) Time-density curve construction and derivation of angiography parametric imaging variables. (C) Parametric mapping: postoperative generation maps of CBF, CBV, MTT, and TTP in a representative internal carotid artery aneurysm with branch incorporation. (D) Region of interests of the reference segment (red), aneurysm (yellow), and branch (green). 3D-DSA, 3-dimensional digital subtraction angiography; An, aneurysm; CBF, cerebral blood flow; CBV, cerebral blood volume; DSA, digital subtraction angiography; MTT, mean transit time; Ref, reference segment; TTP, time to peak.

Hemodynamic metrics

Hemodynamic metrics derived from TDCs were standardized against a reference segment to compensate for variations in contrast agent injection speed and volume across different vascular territories. Three regions of interest (ROIs) were defined for each aneurysm: (I) the reference segment (Ref)—the site where the contrast agent was injected, such as the petrous segment of the internal carotid artery or the V4 segment of the vertebral artery; (II) the branch segment—the artery originating from the aneurysm dome or neck; and (III) the aneurysm segment—the aneurysm sac (9).

From the smoothed TDCs, the following hemodynamic parameters were extracted for both the aneurysm segments: (I) aneurysm cerebral blood volume (CBV)—the area under the aneurysm TDC normalized to the Ref value; (II) aneurysm time to peak (TTP)—the interval to reach maximum intensity in the aneurysm segment normalized to the Ref; (III) aneurysm mean transit time (MTT)—the time during which intensity values exceeded half of the peak value in the aneurysm segment normalized to the Ref; and (IV) aneurysm cerebral blood flow (CBF)—the ratio of the CBV to MTT for the aneurysm normalized to the Ref.

Similarly, the following parameters were calculated for the branch segment: (I) branch CBV—the area under the TDC of the branch artery segment normalized to the Ref; (II) branch TTP—the time to reach the peak intensity of the branch artery segment normalized to the Ref; (III) branch MTT—the duration for which intensity values exceeded half of the peak value in the branch artery segment normalized to the Ref; and (IV) branch CBF—the ratio of the CBV to MTT in the branch artery segment normalized to the Ref.

We calculated the differences between aneurysm and branch artery values as follows: (I) ∆CBF = aneurysm CBF – branch CBF; (II) ∆CBV = aneurysm CBV – branch CBV; (III) ∆MTT = aneurysm MTT – branch MTT; and (IV) ∆TTP = aneurysm TTP – branch TTP.

Statistical analysis

Statistical analysis was performed with SPSS 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as the median and interquartile range, while categorical variables are expressed as frequencies and percentages. Initially, univariate logistic analysis was performed on all parameters. Those showing significant associations were subsequently included in a multivariate logistic model, which was constructed with a stepwise forward selection approach to determine independent risk factors. Receiver operating characteristic (ROC) analysis was conducted to evaluate the discriminative ability of significant variables, with the area under the curve (AUC) and optimal cutoff values being reported. Ninety-five percent confidence intervals (CIs) for AUC were obtained via stratified bootstrapping with 1,000 resamples. A P value below 0.05 was considered statistically significant. In the subgroup analysis, the Benjamini-Hochberg (BH) correction procedure was used to control the false-discovery rate, with adjusted P (P-adj) values <0.05 deemed significant. All continuous variables were rounded to two decimal places.

Results

Demographics and aneurysm-related characteristics

A total of 111 patients with 113 IAs incorporating branches were included in the study. Based on DSA follow-up, 52 IAs (46.02%) were classified as incomplete occlusion, while 61 IAs (54.0%) were classified as complete occlusion. A total of 113 branches were analyzed, with the posterior inferior cerebellar artery (n=33; 29.20%) and posterior communicating artery (Pcom) (n=30; 26.55%) being the most frequently involved branches. The most commonly involved artery was the ophthalmic artery (n=25, 22.12%), followed by middle cerebral artery (n=10, 8.85%), the anterior cerebral artery (ACA) (n=5, 4.42%), the anterior inferior cerebellar artery (n=4, 3.54%), the vertebral artery (n=2, 1.77%), the superior cerebellar artery (n=21.77%), and the posterior cerebral artery (PCA) (n=2, 1.77%).

Univariate analysis revealed notable disparities between the incomplete and complete occlusion groups (Table 1). The incomplete occlusion group, as compared to the complete occlusion group, was more likely to be male (59.62% vs. 31.15%; P=0.003) and have a history of smoking (32.69% vs. 13.11%; P=0.015). Among the aneurysm-related characteristics, posterior circulation aneurysms were significantly more prevalent in the incomplete occlusion group (53.85% vs. 24.59%; P=0.002). The incidence of saccular aneurysms was markedly reduced in the incomplete occlusion group relative to the complete occlusion group (48.08% vs. 75.41%; P=0.003). Notably, among the 43 posterior circulation aneurysms, 9 (20.93%) were saccular; however, no statistically significant difference in their distribution was observed between the incomplete and complete occlusion groups (55.56% vs. 44.44%; P=0.501). Additionally, treatment of FD plus coiling exhibited a higher frequency in the complete occlusion group than in the incomplete occlusion group (52.46% vs. 23.08%; P=0.002). Among the 44 cases involving adjunctive coiling, the median number of coils used was 3.00, with no significant difference between the incomplete and complete occlusion groups (both medians 3.00; P=0.218). For branch characteristics, the incomplete occlusion group, as compared to the complete occlusion group, also had larger branch diameters (1.32 vs. 1.02 mm; P=0.005), although there were no significant differences in whether the branch originated from the dome or neck of the aneurysm (36.54% vs. 27.87%; P>0.05). Additionally, the incomplete occlusion group had a longer median follow-up duration (18.0 vs. 13.0 months; P=0.393), but this was not statistically significant.

Hemodynamic parameters

Univariate analysis of hemodynamic parameters revealed that aneurysm CBF values were significantly elevated in the incomplete occlusion group relative to the complete occlusion group (1.03 vs. 0.74; P<0.001); additionally, aneurysm CBV (1.01 vs. 0.87; P=0.005), aneurysm MTT (0.89 vs. 0.65; P=0.004), and aneurysm TTP (1.35 vs. 1.58; P=0.022) were significantly different between the two groups. Although branch CBF, CBV, and TTP were not significantly different between the groups (P>0.05), branch MTT was significantly elevated in the incomplete occlusion group (0.82 vs. 0.62; P=0.006). Differences in hemodynamic parameters between aneurysm and branch segments (ΔCBF, ΔCBV, ΔMTT, and ΔTTP) were not statistically significant (P>0.05).

Multivariate analysis and predictive performance

Under stepwise forward methods, multivariate logistic regression identified three independent predictors of incomplete occlusion: posterior circulation aneurysm [odds ratio (OR) =2.729; 95% confidence interval (CI): 1.153–6.460; P=0.022], branch diameter (OR =2.802; 95% CI: 1.288–6.097; P=0.009), and aneurysm CBF (OR =10.829; 95% CI: 2.243–52.273; P=0.003). The ROC curve indicated that the combined models incorporating aneurysm CBF, branch diameter, and posterior circulation aneurysm had the highest predictive performance, with an AUC of 0.766 (95% CI: 0.670–0.849). Individually, aneurysm CBF had an AUC of 0.695 (95% CI: 0.590–0.789), branch diameter had an AUC of 0.678 (95% CI: 0.578–0.764), and posterior circulation aneurysm had an AUC of 0.646 (95% CI: 0.577–0.734) (Figure 2). The cutoff value of aneurysm CBF was 1.07.

Figure 2 Receiver operating characteristic curves with 95% CIs for predicting risk factors of incomplete aneurysm occlusion after flow diverter treatment. The area under the curves for the combined model, aneurysm CBF, branch diameter, and posterior circulation aneurysms were 0.766 (95% CI: 0.670–0.849), 0.695 (95% CI: 0.590–0.789), 0.678 (95% CI: 0.578–0.764), and 0.646 (95% CI: 0.557–0.734), respectively. AUC, area under the curve; CBF, cerebral blood flow; CI, confidence interval.

Subgroup analysis: FD alone vs. FD plus coiling

A subgroup analysis was conducted to evaluate the impact of additional coiling on occlusion outcomes (Table 2). This analysis included two groups: the FD-alone group and the FD-plus-coiling group. In the FD-alone group, posterior circulation aneurysm was more frequently associated with incomplete occlusion (P=0.032; P-adj =0.107), as were larger branch diameters (P=0.018; P-adj =0.090) and longer branch MTT (P=0.011; P-adj =0.110). In contrast, in the FD-plus-coiling subgroup, only aneurysm CBF (P=0.046; P-adj =0.460) was significant. However, after BH correction was applied for multiple comparisons, none of these variables retained statistical significance (all P-adj values >0.05), underscoring the potential impact of the limited sample size on statistical power.

Discussion

Several studies have highlighted the impact of incorporated branches on aneurysm occlusion outcomes, suggesting the need for further investigation into the mechanisms by which these branches may hinder complete occlusion (16,17). It has been suggested that the presence of a pressure gradient between the aneurysm and its incorporated branch can influence occlusion outcomes by generating an aspiration effect that sustains residual flow even after FD placement (18). Therefore, our study provides novel insights into the hemodynamic and morphological determinants of incomplete occlusion in IAs with branch incorporation following FD treatment. By integrating clinical, morphological, and QDSA-derived variables, we identified posterior circulation aneurysm, branch diameter, and aneurysm CBF as independent predictors of incomplete occlusion, and the combined models demonstrated better performance did than any single variable. Additionally, subgroup analyses revealed distinct hemodynamic interactions between the FD-alone and FD-plus-coiling strategies.

Posterior circulation aneurysm

Previous studies have consistently recognized that patients with posterior circulation aneurysms tend to have high rates (21.9–25%) of incomplete occlusion after 1 year (19,20). Due to the involvement of multiple perforating arteries, morphological variability (e.g., dissecting, fusiform, and dolichoectatic aneurysms), and complex anatomical structures, posterior circulation aneurysms are associated with higher complication rates and lower occlusion rates (21,22). A meta-analysis on aneurysms with branch incorporation found that posterior circulation aneurysms, such as those involving the posterior inferior cerebellar artery, had an occlusion rate of only 59.1% (23). Furthermore, posterior circulation branch arteries show lower occlusion rates and higher complication risks as compared to anterior circulation branches such as the ACA and Pcom (23). Our findings further demonstrated that among aneurysms incorporating branches, those located in the posterior circulation were 2.729 times more likely to remain incompletely occluded as compared to anterior circulation aneurysms. This increased risk may be due to anatomical complexity and enhanced collateral circulation (5,24). Wu et al. used computational fluid dynamics simulations to demonstrate that persistent blood flow in posterior circulation aneurysms is not fully interrupted after FD placement, with some cases showing increased branch artery flow velocity (25).

Branch diameter

We found that the odds of incomplete occlusion increased by 2.802 fold for each 1 mm increase in branch diameter. The branch diameter reflects the metabolic demand of the tissue it supplies. For example, the definition of an embryonic Pcom corresponds to a PCA-P2 diameter, with a normal diameter of approximately 2.3 mm (range, 1.3–3.1 mm) (26), which is close to the median value of 1.32 mm observed in the incomplete occlusion group in our study. Previous studies have suggested that the presence of an embryonic Pcom can affect the occlusion rate of aneurysms treated with FD and that coiling be considered in cases with larger branches (5). In our study, the influence of branches on aneurysm occlusion was demonstrated by the prolonged MTT in both the aneurysms and branches within the incomplete occlusion group. Previous research by You et al. indicated that reduced MTT after FD placement signifies impaired blood flow, typically manifesting as branch narrowing or occlusion (9). In our study, elevated branch MTT in the incomplete occlusion group was associated with persistent branch flow, which maintained aneurysm perfusion at a reduced velocity, thus hindering complete occlusion. It is plausible that the interplay between aneurysm and branch MTT explains why branch parameters were not independent risk factors in the multivariate analysis. In terms of physiology and hemodynamics, branch MTT was significant in the univariate analysis, reflecting the degree of branch flow impairment, which aligns with the findings by You et al., who reported that the tissue demands an adequate blood supply (9).

Aneurysm CBF

Aneurysm CBF is the immediate flow rate per unit time after the procedure, normalized against the initial inflow of contrast agent (14). Reduced CBF may facilitate aneurysm occlusion after FD treatment. Our results align with the research by Cebral et al., which indicated that FDs induce thrombosis through decreased flow velocity, thereby encouraging fibrin accumulation within the aneurysm (27). Notably, compared to the other two independent risk factors, aneurysm CBF can be dynamically adjusted during surgery in time. The cutoff value of aneurysm CBF was 1.07. The combined model incorporating all three predictors showed the best predictive performance, suggesting that surgeons can optimize treatment strategies to lower aneurysm CBF and promote higher occlusion rates. Additionally, the hemodynamic differences between the aneurysm and its branches were not significantly associated with incomplete occlusion. However, previous studies have suggested that the presence of a pressure gradient is often closely linked to the existence of branch vessels after FD deployment (5,18). This may indicate that collateral circulation plays a compensatory role over time. For instance, in aneurysms involving the ophthalmic artery, sufficient collateral flow from the external carotid artery can lead to gradual, asymptomatic occlusion of the ophthalmic artery, ultimately resulting in favorable aneurysm healing (5,18,28). These observations underscore the importance of assessing both aneurysm location and the adequacy of collateral circulation (18). The complex interplay of hemodynamic factors may explain why pressure gradients were not statistically significant in our analysis, suggesting that the tissue’s physiological demand for branch perfusion may play a more pivotal role in determining aneurysm occlusion outcomes.

Subgroup analysis: the role of coiling in FD treatment

The differential hemodynamic profiles between the FD-alone and FD-coiling subgroups support existing hypotheses regarding the mechanism of the coiling. In the FD-alone subgroup, the association of branch MTT with incomplete occlusion implies prolonged branch hemodynamic changes. In contrast, in the FD-plus-coiling subgroup, the isolated CBF significance suggests that coiling primarily focuses the blood flow on the aneurysm itself, reducing the impact of branch blood flow on the aneurysm’s healing status, with branch metrics becoming less significant. Neurointerventionalists typically use adjunctive coiling for aneurysms with larger incorporated branches, which explains why across the entire cohort and within subgroups, branch and aneurysm MTTs were significantly higher in the incomplete occlusion group and why branch MTT was comparatively lower in the incomplete occlusion subgroup treated with FD alone. The interplay between branch and aneurysm MTT may partly account for their lack of significance in multivariate analysis. In FD-alone subgroup, although posterior circulation aneurysms (P=0.032) and larger branch diameters (P=0.018) initially showed statistical significance, these associations did not remain significant after correction for multiple comparisons (P-adj >0.05). Nevertheless, the observed tendencies suggest that FD alone may be less effective for aneurysms involving larger branches or located in the posterior circulation. In the FD-plus-coiling subgroup, these variables were not statistically significant, implying that adjunctive coiling may facilitate aneurysm occlusion, in line with previous findings (29). However, a recent study indicated that dense coil packing may increase the risk of ischemic complications, underscoring the need for caution in coiling procedures (13). After BH correction was applied, none of the metrics reached statistical significance, likely due to the limited sample size, yet a clear statistical tendency remained. Future studies with larger, real-time prospective cohorts are needed to further investigate the impact of coil embolization on incomplete occlusion outcomes.

Limitations

There are several limitations to consider in relation to this study. First, the retrospective design and single-center data source may reduce the generalizability of our findings. Second, due to the sample size, we were unable to perform a more detailed subgroup analysis of aneurysm locations and involvement of specific branches, such as the vertebrobasilar artery, Pcom, and ACA. Third, the limited sample size and number of incomplete occlusion events may reduce the power and stability of the multivariate analysis. Despite using stepwise forward selection to avoid overfitting, the events per variable for some predictors was below the recommended threshold of 10, which might have introduced bias and reduced coefficient precision. Fourth, most of the coiled cases were loosely packed, and coil-related artifacts may have an effect on parameter calculation; nonetheless, the overall findings remain robust. Only the number of coils was reported, and packing density was not calculated due to the small sample size and the complexity of precise assessment. Finally, although subgroup analysis suggested that reducing aneurysm CBF could promote aneurysm occlusion, dense coiling may increase the risk of complications, and we did not provide a defined safety threshold. Prospective studies with larger sample sizes should be conducted to further refine these findings.

Conclusions

Our study identified three independent risk factors for incomplete occlusion in FD-treated aneurysms with branch incorporation: posterior circulation aneurysms, larger branch diameters, and higher aneurysm CBF. These findings highlight the importance of both morphological and hemodynamic factors being considered in the assessment of incomplete occlusion and clinical decision-making.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-358/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-358/dss

Funding: This study was supported by the National Natural Science Foundation of China (grant No. 82372058); China Postdoctoral Science Foundation (grant No. 2024M75178); and Postdoctoral Fellowship Program of CPSF (grant No. GZC20241099).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-358/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 retrospective study was approved by the Institutional Review Board of Beijing Tiantan Hospital (approval ID: KY2023-261-01). Due to the retrospective nature of the study, written informed consent from participants was waived. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

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. Lee RP, Bhimreddy M, Kim J, Wicks RT, Xu R, Bender M, Yang W, Sattari SA, Hung A, Jackson CM, Gonzalez LF, Huang J, Tamargo R, McDougall CG, Caplan JM. No Delayed Ruptures on Long-Term Follow-Up of a Case Series of Persistently Filling Saccular Internal Carotid Artery Aneurysms After Flow Diversion With the Pipeline Embolization Device. Neurosurgery 2023;93:994-9. [Crossref] [PubMed]
2. Dong L, Wang C, Chen X, Li M, Li T, Liu H, Zhao Y, Duan R, Jin W, Zhang Y, Wang Y, Lv M. Predicting Persistent Aneurysm Filling After Pipeline Embolization Device Treatment in Patients with Intracranial Aneurysm: Development and External Validation of a Nomogram Model. Transl Stroke Res 2025;16:392-402. [Crossref] [PubMed]
3. Dandapat S, Mendez-Ruiz A, Martínez-Galdámez M, Macho J, Derakhshani S, Foa Torres G, Pereira VM, Arat A, Wakhloo AK, Ortega-Gutierrez S. Review of current intracranial aneurysm flow diversion technology and clinical use. J Neurointerv Surg 2021;13:54-62. [Crossref] [PubMed]
4. Moshayedi H, Omofoye OA, Yap E, Oyekunle TO, Sasaki-Adams DM, Solander SY. Factors Affecting the Obliteration Rate of Intracranial Aneurysms Treated with a Single Pipeline Embolization Device. World Neurosurg 2017;104:205-12. [Crossref] [PubMed]
5. Daou B, Valle-Giler EP, Chalouhi N, Starke RM, Tjoumakaris S, Hasan D, Rosenwasser RH, Hebert R, Jabbour P. Patency of the posterior communicating artery following treatment with the Pipeline Embolization Device. J Neurosurg 2017;126:564-9. [Crossref] [PubMed]
6. Stahl J, Marsh LMM, Thormann M, Ding A, Saalfeld S, Behme D, Berg P. Assessment of the flow-diverter efficacy for intracranial aneurysm treatment considering pre- and post-interventional hemodynamics. Comput Biol Med 2023;156:106720. [Crossref] [PubMed]
7. Zhang M, Tian Z, Zhang Y, Zhang Y, Wang K, Leng X, Yang X, Xiang J, Liu J. How to perform intra-aneurysmal coil embolization after Pipeline deployment: a study from a hemodynamic viewpoint. J Neurointerv Surg 2023;15:157-62. [Crossref] [PubMed]
8. Shiraz Bhurwani MM, Waqas M, Podgorsak AR, Williams KA, Davies JM, Snyder K, Levy E, Siddiqui A, Ionita CN. Feasibility study for use of angiographic parametric imaging and deep neural networks for intracranial aneurysm occlusion prediction. J Neurointerv Surg 2020;12:714-9. [Crossref] [PubMed]
9. You W, Wei D, Gui S, Jiang J, Chen T, Tang Y, Ye W, Lv J, Lin J, Chen P, Wang Z, Gong W, Jin H, Ge H, Jiang Y, Sun Y, Li Y. Quantitative Analysis of Hemodynamic Changes in Branch Arteries Covered by Flow Diverters. Neurosurgery 2024;95:1167-77. [Crossref] [PubMed]
10. Gölitz P, Struffert T, Rösch J, Ganslandt O, Knossalla F, Doerfler A. Cerebral aneurysm treatment using flow-diverting stents: in-vivo visualization of flow alterations by parametric colour coding to predict aneurysmal occlusion: preliminary results. Eur Radiol 2015;25:428-35. [Crossref] [PubMed]
11. Struffert T, Ott S, Kowarschik M, Bender F, Adamek E, Engelhorn T, Gölitz P, Lang S, Strother CM, Doerfler A. Measurement of quantifiable parameters by time-density curves in the elastase-induced aneurysm model: first results in the comparison of a flow diverter and a conventional aneurysm stent. Eur Radiol 2013;23:521-7. [Crossref] [PubMed]
12. Liang F, Ma C, Zhu H, Liu L, Liang S, Jiang P, Zhang Y, Jiang C. Using angiographic parametric imaging-derived radiomics features to predict complications and embolization outcomes of intracranial aneurysms treated by pipeline embolization devices. J Neurointerv Surg 2022;14:826-31. [Crossref] [PubMed]
13. Tong X, Han M, Xue X, Wu Z, Chen J, Liu A. Coiling embolization strategy for medium-to-giant-sized intracranial aneurysms treated with pipeline embolization device: a propensity score-weighted study. Eur Radiol 2023;33:7967-77. [Crossref] [PubMed]
14. Hu SQ, Turhon M, Liu T, Li W, Yang X. Machine Learning-Assisted Prediction of Persistent Incomplete Occlusion in Intracranial Aneurysms From Angiographic Parametric Imaging-Derived Features. Acad Radiol 2025;32:5425-34. [Crossref] [PubMed]
15. Zhang Y, Zhu H, Cao T, Zhang L, Chang Y, Liang S, Ma C, Liang F, Song Y, Zhang J, Li C, Jiang C. Rupture-Related Features of Cerebral Arteriovenous Malformations and Their Utility in Predicting Hemorrhage. Stroke 2024;55:1339-48. [Crossref] [PubMed]
16. Hammoud B, El Zini J, Awad M, Sweid A, Tjoumakaris S, Jabbour P. Predicting incomplete occlusion of intracranial aneurysms treated with flow diverters using machine learning models. J Neurosurg 2024;140:1716-25. [Crossref] [PubMed]
17. Fujitani S, Tsuruta W, Tomioka A, Ishigami D, Sekine T, Hosoo H, Kamiya Y, Matsumaru Y. Aneurysm Isolation is Associated with Complete Occlusion of Aneurysms After Flow Diverter Treatment. Clin Neuroradiol 2023;33:1087-93. [Crossref] [PubMed]
18. Griessenauer CJ, Ogilvy CS, Foreman PM, Chua MH, Harrigan MR, Stapleton CJ, Patel AB, He L, Fusco MR, Mocco J, Winkler PA, Patel AS, Thomas AJ. Pipeline Embolization Device for small paraophthalmic artery aneurysms with an emphasis on the anatomical relationship of ophthalmic artery origin and aneurysm. J Neurosurg 2016;125:1352-9. [Crossref] [PubMed]
19. Qi P, Tong X, Liang X, Xue X, Wu Z, Feng X, Zhang M, Jiang Z, Wang D, Liu A. Flow diversion for posterior circulation aneurysms: a multicenter retrospective study. Ther Adv Neurol Disord 2023;16:17562864231176187. [Crossref] [PubMed]
20. Alwakeal A, Shlobin NA, Golnari P, Metcalf-Doetsch W, Nazari P, Ansari SA, Hurley MC, Cantrell DR, Shaibani A, Jahromi BS, Potts MB. Flow Diversion of Posterior Circulation Aneurysms: Systematic Review of Disaggregated Individual Patient Data. AJNR Am J Neuroradiol 2021;42:1827-33. [Crossref] [PubMed]
21. Fang YB, Wen WL, Yang PF, Zhou Y, Wu YN, Hong B, Xu Y, Zhao WY, Liu JM, Huang QH. Long-Term Outcome of Tubridge Flow Diverter(S) in Treating Large Vertebral Artery Dissecting Aneurysms-A Pilot Study. Clin Neuroradiol 2017;27:345-50. [Crossref] [PubMed]
22. Kiyofuji S, Graffeo CS, Perry A, Murad MH, Flemming KD, Lanzino G, Rangel-Castilla L, Brinjikji W. Meta-analysis of treatment outcomes of posterior circulation non-saccular aneurysms by flow diverters. J Neurointerv Surg 2018;10:493-9. [Crossref] [PubMed]
23. Liu J, Cao F, Zhenmei N, Guo Y, Li Y, Yuan D, Jiang W, Yan J. Flow-diverter stents in intracranial aneurysm treatment: impact on covered cerebral artery branches. Int J Surg 2024;110:53-65. [Crossref] [PubMed]
24. Roy AK, Howard BM, Haussen DC, Osbun JW, Halani SH, Skukalek SL, Tong F, Nogueira RG, Dion JE, Cawley CM, Grossberg JA. Reduced Efficacy of the Pipeline Embolization Device in the Treatment of Posterior Communicating Region Aneurysms with Fetal Posterior Cerebral Artery Configuration. Neurosurgery 2018;82:695-700. [Crossref] [PubMed]
25. Wu X, Tian Z, Liu J, Zhang Y, Li W, Zhang Y, Chen J, Zhou Y, Yang X, Mu S. Patency of Posterior Circulation Branches Covered by Flow Diverter Device: A Hemodynamic Study. Front Neurol 2019;10:658. [Crossref] [PubMed]
26. Uz A. The segmentation of the posterior cerebral artery: a microsurgical anatomic study. Neurosurg Rev 2019;42:155-61. [Crossref] [PubMed]
27. Cebral JR, Mut F, Löhner R, Marsh L, Chitsaz A, Bilgin C, Bayraktar E, Kallmes D, Kadirvel R. Influence of Vessel Geometry, Flow Conditions, and Thrombin Concentration on Fibrin Accumulation and Cerebral Aneurysm Occlusion After Flow Diversion. Int J Numer Method Biomed Eng 2025;41:e3904. [Crossref] [PubMed]
28. Fiorella D, Lylyk P, Szikora I, Kelly ME, Albuquerque FC, McDougall CG, Nelson PK. Curative cerebrovascular reconstruction with the Pipeline embolization device: the emergence of definitive endovascular therapy for intracranial aneurysms. J Neurointerv Surg 2018;10:i9-i18. [Crossref] [PubMed]
29. Kang H, Luo B, Liu J, Zhang H, Li T, Song D, Zhao Y, Guan S, Maimaitili A, Wang Y, Feng W, Wang Y, Wan J, Mao G, Shi H, Zhang Y, Yang X. Postoperative occlusion degree after flow-diverter placement with adjunctive coiling: analysis of complications. J Neurointerv Surg 2022;14:371-5. [Crossref] [PubMed]