Advertisement

Risk factors for the rupture of mirror middle cerebral artery aneurysm using computer-assisted semiautomated measurement and hemodynamic analysis

Open AccessPublished:November 02, 2022DOI:https://doi.org/10.1016/j.jstrokecerebrovasdis.2022.106841

      Highlights

      • Mirror MCA aneurysms are rare but could avoid the patient-related factors and selection bias.
      • We use computer-assisted semiautomated measurement (CASAM) to measure the morphologic parameters more precisely. We also used computational fluid dynamics (CFD).
      • The size, neck diameter and maximum oscillatory shear index was independently significant and emphasized in the study.

      Abstract

      Objectives

      To identify the morphologic and hemodynamic risk factor of mirror middle cerebral artery (MCA) aneurysms.

      Methods

      We conducted a retrospective analysis of 40 paired mirror MCA aneurysms. Aneurysms were divided into ruptured and unruptured groups. Seventeen morphological and nine hemodynamic parameters were measured using computer-assisted semiautomated measurement (CASAM) and computer flow dynamic (CFD) simulation. We performed a paired t-test (for normally distributed data) or a paired Wilcoxon rank-sum (for non-normally distributed data) to analyze all parameters between the groups. Multivariate conditional logistic regression analysis identified independent risk factors. The receiver operating characteristic curve was analyzed to acquire the area under the curve (AUC) and the cutoff values of the independent risk factors.

      Results

      There were significant differences in morphological and hemodynamic parameters between the ruptured and unruptured mirror aneurysms. The multivariate logistic analysis showed that the greater size (odds ratio [OR] = 9.807, p = 0.003), smaller neck diameter (OR = 0.285, p = 0.018) and maximum oscillatory shear index (OSI) (OR = 0.000001, p = 0.046) were independently correlated with aneurysm rupture. AUCs for size, N. and maximum OSI were 0.794, 0.695, and 0.701, respectively. The cutoff values of the size, neck diameter, and maximum OSI were 6.30, 5.07, and 0.356437, respectively.

      Conclusions

      Morphology and hemodynamics can help predict aneurysm rupture risks. The more significant size, smaller neck diameter and maximum OSI were independent risk factors for the rupture of MCA aneurysms. The variables could aid practical risk evaluation.

      Keywords

      Introduction

      Intracranial aneurysms (IAs) are associated with high rates of death and disability.
      • Neifert S.N.
      • Chapman E.K.
      • Martini M.L.
      • et al.
      Aneurysmal subarachnoid hemorrhage: the last decade.
      ,
      • Xu W.D.
      • Chen R.D.
      • Hu S.Q.
      • et al.
      Morphological evaluation of the risk of posterior communicating artery aneurysm rupture: a mirror aneurysm model.
      Their rupture is accompanied by subarachnoid hemorrhage (SAH) and other severe complications.
      • Neifert S.N.
      • Chapman E.K.
      • Martini M.L.
      • et al.
      Aneurysmal subarachnoid hemorrhage: the last decade.
      ,
      • Xu W.D.
      • Wang H.
      • Wu Q.
      • et al.
      Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
      Predicting the aneurysms that are prone to rupture is crucial. A study showed that IAs are generated by morphologic changes secondary to blood flow-induced vessel remodeling.
      • Frösen J.
      • Cebral J.
      • Robertson A.M.
      • et al.
      Flow-induced, inflammation-mediated arterial wall remodeling in the formation and progression of intracranial aneurysms.
      Hemodynamic and morphological numerically quantified rupture risk.
      • Diagbouga M.R.
      • Morel S.
      • Bijlenga P.
      • et al.
      Role of hemodynamics in initiation/growth of intracranial aneurysms.
      • Soldozy S.
      • Norat P.
      • Elsarrag M.
      • et al.
      The biophysical role of hemodynamics in the pathogenesis of cerebral aneurysm formation and rupture.
      • Han P.
      • Jin D.
      • Wei W.
      • et al.
      The prognostic effects of hemodynamic parameters on rupture of intracranial aneurysm: a systematic review and meta-analysis.
      However, the risks of rupture are also influenced by patient-related factors such as race, gender, hypertension, and smoking. Mirror aneurysms could avoid patient-related issues by exploiting pair analysis.
      • Xu L.
      • Wang H.
      • Chen Y.
      • et al.
      Morphological and hemodynamic factors associated with ruptured middle cerebral artery mirror aneurysms: a retrospective study.
      Studies focusing on mirror aneurysms returned inconsistent results. Xu et al. demonstrated that width, aspect ratio (AR), and size ratio (SR) were independently correlated with the rupture.
      • Xu W.D.
      • Wang H.
      • Wu Q.
      • et al.
      Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
      However, Maslehaty et al. identified only the size of the dome as an independent risk factor.
      • Maslehaty H.
      • Capone C.
      • Frantsev R.
      • et al.
      Predictive anatomical factors for rupture in middle cerebral artery mirror bifurcation aneurysms.
      Hemodynamics simulated using computational fluid dynamics (CFD) can be used to analyze flow-induced changes in aneurysms.
      • Hu S.Q.
      • Chen R.D.
      • Xu W.D.
      • et al.
      A predictive hemodynamic model based on risk factors for ruptured mirror aneurysms.
      The role of hemodynamics remains unclear and needs to be further explored.
      • Murayama Y.
      • Fujimura S.
      • Suzuki T.
      • et al.
      Computational fluid dynamics as a risk assessment tool for aneurysm rupture.
      Aneurysm rupture risk is location-dependent, and the inclusion of mirror aneurysms in different sites returns confusing results.
      • Tang X.
      • Zhou L.
      • Wen L.
      • et al.
      Morphological and hemodynamic characteristics associated with the rupture of multiple intracranial aneurysms.
      ,
      • Liu J.
      • Chen Y.
      • Zhu D.
      • et al.
      A nomogram to predict rupture risk of middle cerebral artery aneurysm.
      Therefore, we measured the morphology and hemodynamics in mirror middle cerebral artery (MCA) aneurysms to identify relevant parameters for rupture risk.

      Methods

      Patients and data

      Our hospital's institutional ethics committee approved the study, and we collected the data following the consent of the patients or their close relatives We retrospectively reviewed the digital subtraction angiograms (DSAs) of IAs from October 2012 to December 2021, including 432 MCA aneurysms. Of these, 412 MCA aneurysms could be reconstructed using computer-assisted semiautomated measurement (CASAM); 39 aneurysms were excluded for the following reasons: (a) fusiform or dissecting aneurysm (n = 11); (b) presence of arteriovenous malformations or moyamoya disease (MMD) (n = 9); (c) extensive hemorrhage imaging in which it was challenging to identify the ruptured side (n = 9); (d) history of aneurysm treatment (n = 10). In 373 saccular MCA aneurysms, 40 paired mirror MCA aneurysms were included for the following reasons: (a) saccular shape; (b) bilateral MCA bifurcation aneurysms; (c) evident SAH on imaging; (d) one MCA aneurysm ruptured. These 40 paired mirror MCA aneurysms were divided into a ruptured group (n = 40) and an unruptured group (n = 40). We excluded 293 saccular MCA aneurysms: (a) unruptured mirror MCA aneurysms (n = 7) and (b) single MCA aneurysms (n = 279).
      Patients with SAH underwent emergent CT on admission and DSA within 72 h after hemorrhage. Of the 40 paired mirror MCA aneurysms, all ruptured aneurysms were treated by endovascular therapy or surgery. Fourteen unruptured aneurysms underwent endovascular therapy, and sixteen underwent surgery. Eight patients were followed up yearly for unruptured aneurysms. Two patients were lost to follow-up.

      Morphologic variable measurement

      Pretreatment DSA images and three-dimensional (3D) reconstructions for each patient were obtained using the Innova Workplace (GE Medical). CASAM was used to measure the morphologic parameters. The mechanism of CASAM has been described in detail.
      • Geng J.
      • Hu P.
      • Ji Z.
      • et al.
      Accuracy and reliability of computer-assisted semi-automated morphological analysis of intracranial aneurysms: an experimental study with digital phantoms and clinical aneurysm cases.
      CASAM procedures are shown in Fig. 1: (a) importing of original DSA DICOM data (Fig. 1A); (b) 3D vessel reconstruction (Fig. 1B); (c) manual selection of the region of interest, including aneurysm and parent artery (Fig. 1C); (d) automated extraction of the aneurysm (Figure 1D); (e) automated measurement and exportation. Morphologic parameter definitions are presented in Fig. 1 E-F. Results were accurate to two decimal places. The definitions of morphologic variables are as follows:
      Fig 1
      Fig. 1Procedures of Computer-assisted semiautomated measurement (CASAM). (A) Importing original DSA DICOM; (B) 3D vessel reconstruction; (C) manual selection of the region of interest (ROI), including aneurysm and parent artery; (D) automated extraction of the aneurysm (blue part); (E-F) Definition of relevant morphological parameters. D, diameter; WD, width; H, height; WH, aneurysm width; N, neck diameter; IA, incidence angle; AA, aneurysm angle. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
      Diameter (D): The maximum distance from the center of the neck to the top.
      Width (WD): The maximum distance perpendicular to the aneurysm diameter (D).
      Height (H): The maximum distance of the dome from the center of the neck plane.
      Aneurysm width (WH): The maximum diameter perpendicular to the height.
      Size: The maximum diameter of the aneurysm.
      Neck diameter (N): The maximum diameter in the neck plane.
      Volume (V): Aneurysm lumen volume calculated in 3D space.
      Neck space (S): The area of the projection plane of the neck surface.
      Incident angle (IA): The angle between the aneurysm diameter and the centerline of the parent artery.
      Aneurysm angle (AA): The angle between the aneurysm diameter and the projection plane of the neck surface.
      D/WD ratio: The ratio of diameter to width.
      H/WH ratio: The ratio of height to aneurysm width.
      Bottleneck factor (BNF): width/neck width.
      Undulation index (UI): UI=1VVCH where VCH is the convex hull volume of the aneurysm.
      Non-sphericity index (NSI): NSI=1(18π)13*V23S
      SR: the ratio of diameter to the parent artery diameter
      AR: the ratio of height to neck width

      Hemodynamic variable calculation

      CFD simulations are presented in Fig. 1 (A–B). The open-source CFD software package OpenFoam was used for grid generation and CFD calculation. Then three-dimensional DSA data were imported into the system. Mesh Generation: We OpenFoam to generation volume mesh. We used a mesh size of 0.1 mm for the inlet, outlet and sac parts; We used a mesh size of 0.3 mm for the parent artery and other parts. Three layers of prism elements for wall were used for the CFD simulations. The number of finite-volume grid elements used in this study was approximately 1 million. After mesh generation, the vascular model and blood flow state were generated using the following assumptions: the vascular wall was rigid, and the blood flow was isothermal and laminar (an incompressible Newtonian fluid).
      • Jeong W.
      • Han M.H.
      • Rhee K.
      The hemodynamic alterations induced by the vascular angular deformation in stent-assisted coiling of bifurcation aneurysms.
      The blood flow was approximated using the unsteady Navier-Stokes equations. Blood density was set to 1060 kg/m3, and blood viscosity was set to 0.004 N s/m2. Because the individual-specific blood flow conditions cannot be obtained in real-time, the standard data of the normal population obtained from the literature were used as the input boundary conditions, and the outlet was set at 0 pressure state.
      • Ford M.D.
      • Alperin N.
      • Lee S.H.
      • et al.
      Characterization of volumetric flow rate waveforms in the normal internal carotid and vertebral arteries.
      Finally, the calculation duration was set to last for three cardiac cycles. The results of the third cycle reached a stable state. Each cycle was set with 100 time steps. The results from the third simulated cardiac cycle were collected as output for the final analyses.
      The minimum, maximum, and mean values of eight hemodynamic variables were obtained as follows: wall shear stress (WSS), normalized WSS (NWSS), WSS gradient (WSSG), streamwise WSSG, normalized pressure (NP), oscillatory shear index (OSI), combined hemodynamic parameters (CHP), and relative residence time (RRT). The low shear area (LSA) ratio was reported separately. These definitions have been described in other studies.
      • Diagbouga M.R.
      • Morel S.
      • Bijlenga P.
      • et al.
      Role of hemodynamics in initiation/growth of intracranial aneurysms.
      • Soldozy S.
      • Norat P.
      • Elsarrag M.
      • et al.
      The biophysical role of hemodynamics in the pathogenesis of cerebral aneurysm formation and rupture.
      • Han P.
      • Jin D.
      • Wei W.
      • et al.
      The prognostic effects of hemodynamic parameters on rupture of intracranial aneurysm: a systematic review and meta-analysis.

      Statistical analysis

      Statistical analysis was performed using SPSS 20.0 (IBM Inc, Chicago, IL). Continuous variables were expressed as means ± standard deviation, and categorical variables were expressed as frequencies (percentages). The Kolmogorov-Smirnov test was performed to determine whether the parameter dataset was normally distributed. A paired t-test (for normally distributed data) or paired Wilcoxon rank-sum test (for non-normally distributed data) was used for the continuous variables. Significant morphological variables (D, WD, H, WH, Size, N, BNF, SR, and AR) and hemodynamic variables (minimum NWSS, mean NWSS, minimum WSSG, maximum OSI, maximum CHP, mean CHP, maximum RRT, mean RRT, and LSA ratio) were involved in the conditional multivariate logistic regression using a stepwise conditional method. Independent risk factors were derived, and receiver operating characteristic (ROC) curves were used to acquire the area under the curves (AUCs) to evaluate the performance of the independent risk factors (Size, N, and maximum OSI). Cutoff values were acquired to identify the higher risks. Differences where p < 0.05 were statistically significant.

      Results

      Patients and common conditions

      We included 40 patients with mirror MCA aneurysms and divided 80 aneurysms into a ruptured group (n = 40) and an unruptured group (n = 40); 22 (55.0%) patients were female. Ages ranged from 41 to 77 years, with a mean age of 57.8 years; 27 (67.5%) of the patients had hypertension, and 14 (35.0%) had hyperlipemia. Ten patients (25.0%) had a smoking history, and nine (22.5 %) patients had a drinking history. Five patients (12.5%) had a history of heart disease. None had a family history of SAH or inherited diseases such as Marfan's syndrome. On admission, Hunt and Hess grades were I in four patients (10.0%), II in 23 patients (57.5 %), III in six patients (15.0%), and IV in seven patients (17.5%); 19 (47.5 %) ruptured aneurysms were on the right side.

      Morphologic and hemodynamic analyses

      The values of morphological parameters are displayed in Table 1. Significant differences in morphological parameters are shown in Fig. 2 A-B. In the paired t- or Wilcoxon rank-sum tests, the ruptured group had significantly larger D, WD, H, WH, Size, N, BNF, SR, and AR. On average, H/WH (p = 0.368), D/WD (p = 0.051), V (p = 0.083), S (p = 0.054) and IA (p = 0.468) were larger in the ruptured group, and AA (p = 0.313, UI (p = 0.535), and NSI (p = 0.700) were smaller in the ruptured group; however, none of these differences were significant.
      Table 1Measurement and analysis of morphological variables.
      VariablesRuptured group

      (n = 40)
      Unruptured group

      (n = 40)
      Paired t or Wilcoxon rank-sum

      p value
      Diameter, D, mm6.06 ± 2.423.87 ± 2.31< 0.001
      Width, WD, mm6.57 ± 2.294.71 ± 2.550.001
      Height, H5.23 ± 2.303.39 ± 2.09< 0.001
      Aneurysm width, WH, mm6.84 ± 2.254.96±2.630.001
      H/WH0.77 ± 0.250.68 ± 0.190.368
      D/WD0.93 ± 0.230.81 ± 0.140.051
      Size, mm8.16 ± 2.665.51 ± 2.61< 0.001
      Neck diameter, N, mm6.05 ± 2.004.84 ± 2.360.007
      BNF1.14 ± 0.211.01 ± 0.130.002
      Volume, V, mm3136.61 ± 156.0585.59 ± 223.300.083
      Space, S, mm231.78 ± 21.3822.91 ± 28.420.054
      Incidence Angle, IA,o125.87 ± 36.99121.25 ± 26.040.468
      Aneurysm Angle, AA,o44.09 ± 23.2151.23 ± 22.980.313
      UI0.17 ± 0.060.19 ± 0.100.535
      NSI0.375 ± 0.03450.378 ± 0.0620.700
      Size Ratio, SR3.16 ± 1.521.93 ± 1.45< 0.001
      Aspect Ratio, AR0.90 ± 0.370.69 ± 0.230.049
      Variables showing statistical significance (p < 0.05) are in bold. H/WH, height aneurysm width ratio; D/WD, diameter width ratio; BNF, bottle neck factor; UI, undulation index; NSI, non-sphericity index.
      Fig 2
      Fig. 2Nested box plots showing the comparison of significant parameters between the ruptured and unruptured groups. (A) Distribution of the morphological parameters. (B) Distribution of the ratio of morphological parameters; (C-D) Distribution of significant hemodynamic parameters. D, diameter; WD, width; H, height; WH, aneurysm width; N, neck diameter; H/WH, height aneurysm width ratio; D/WD, diameter width ratio; BNF, bottle neck factor; SR, size ratio; AR, aspect ratio; NWSS, normalized wall shear stress; WSSG, wall shear stress gradient; RRT, relative residence time; OSI, oscillatory shear index; CHP, combined hemodynamic parameters; LSA ratio, low shear area ratio.
      The values of hemodynamic parameters are displayed in Table 2. Significant differences in hemodynamic parameters are shown in Fig. 2 C-D. The CFD simulation of significant parameters is shown in Fig. 3. In the paired t- or Wilcoxon rank-sum tests, minimum NWSS (p = 0.041), mean NWSS (p = 0.022), and minimum WSSG (p = 0.006) were significantly smaller in the ruptured group. Maximum OSI (p = 0.004), maximum CHP (p = 0.010), mean CHP (p = 0.014), maximum RRT (p = 0.047), mean RRT (p = 0.006), and LSA ratio (p = 0.007) were significantly larger in the ruptured group.
      Table 2Measurement and analysis of hemodynamic variables.
      VariablesRuptured group

      (n = 40)
      Unruptured group

      (n = 40)
      Paired t or Wilcoxon rank-sum

      p value
      WSS
      Minimum0.24 ± 0.400.89 ± 3.160.202
      Maximum51.20 ± 55.4649.41 ± 37.880.882
      Mean8.29 ± 8.7911.90 ± 12.100.104
      NWSS
      Minimum0.019 ± 0.0270.061 ± 0.120.041
      Maximum4.38 ± 1.904.57 ± 2.090.698
      Mean0.69 ± 0.320.95 ± 0.520.022
      WSSG
      Minimum9.53 ± 14.8325.97 ± 35.480.006
      Maximum11735.40 ± 13394.7110939.49 ± 9424.890.714
      Mean1077.14 ± 1227.581502.78 ± 1434.390.139
      Streamwise WSSG
      Minimum-10345.71 ± 122992.10-9201.61 ± 8280.180.594
      Maximum6500.43 ± 7163.356026.63 ± 2067.430.684
      Mean-178.45 ± 233.88-230.23 ± 440.250.526
      NP
      Minimum0.69 ± 0.340.71 ± 0.350.979
      Maximum1.74 ± 0.521.83 ± 0.630.809
      Mean1.10 ± 0.291.22 ± 0.340.143
      OSI
      Minimum0.00049 ± 0.000790.00039 ± 0.000240.457
      Maximum0.41 ± 0.0750.34 ± 0.120.004
      Mean0.046 ± 0.120.018 ± 0.0130.124
      CHP
      Minimum0.002 ± 0.00170.0017 ± 0.00140.374
      Maximum0.89 ± 0.0950.81 ± 0.140.010
      Mean0.15 ± 0.0620.12 ± 0.0490.014
      RRT
      Minimum0.0046 ± 0.00590.0049 ± 0.0110.827
      Maximum67.97 ± 134.1018.82 ± 63.530.047
      Mean0.20 ± 0.300.071 ± 0.0780.006
      LSA Ratio0.13 ± 0.190.054 ± 0.0710.007
      Variables showing statistical significance (p < 0.05) are in bold; WSS, wall shear stress; NWSS, normalized wall shear stress; WSSG, wall shear stress gradient; NP, normalized pressure; OSI, oscillatory stress index; CHP, combined hemodynamic parameters; RRT, relative residence time; LSA ratio, low shear area ratio.
      Fig 3
      Fig. 3The CFD simulation of significant hemodynamic parameters (p < 0.05) in paired t or Wilcoxon rank-sum analysis. NWSS, normalized wall shear stress; WSSG, wall shear stress gradient; RRT, relative residence time; OSI, oscillatory shear index; CHP, combined hemodynamic parameters; LSA ratio, low shear area ratio.
      Significant variables were included in the multivariate conditional forward method to identify relevant parameters in paired t or Wilcoxon rank-sum tests. As shown in Table 3, the size (OR = 9.807, p = 0.003), neck diameter (OR = 0.285, p = 0.018) and maximum OSI (OR = 0.000001, p = 0.046) were independent risk factors for MCA aneurysm rupture. ROC were applied to 40 pairs of mirror aneurysms, shown in Fig. 4. The AUC values of the size, N, and maximum OSI were 0.794, 0.695, and 0.701 respectively. The cutoff values of the size, N, and maximum OSI were 6.30, 5.07, and 0.356437, respectively.
      Table 3Independent risk factors using multivariable logistic regression analysis.
      VariablesβSEWaldp valueOR (95% CI)
      Size, mm2.2830.7768.6560.0039.807 (2.143 – 44.878)
      Neck diameter, mm−1.2550.5335.5540.0180.285 (0.100 – 0.809)
      OSI max−14.0387.0273.9910.0460.000001 (0.001 – 0.766)
      OSI, oscillatory shear index; OR, odds ratio; CI, confidence interval.
      Fig 4
      Fig. 4ROC curve showing the AUCs of independent risk factors in 40 paired mirror MCA aneurysms.

      Discussion

      The advantage of using mirror aneurysm is that it does not require consideration of underlying diseases such as blood pressure or lifestyle-related diseases.
      • Xu W.D.
      • Chen R.D.
      • Hu S.Q.
      • et al.
      Morphological evaluation of the risk of posterior communicating artery aneurysm rupture: a mirror aneurysm model.
      ,
      • Xu W.D.
      • Wang H.
      • Wu Q.
      • et al.
      Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
      However, studies that ignored site-specific mirror multiple aneurysms showed inconsistent results.
      • Tang X.
      • Zhou L.
      • Wen L.
      • et al.
      Morphological and hemodynamic characteristics associated with the rupture of multiple intracranial aneurysms.
      ,
      • Yuan J.
      • Huang C.
      • Lai N.
      • et al.
      Hemodynamic and morphological analysis of mirror aneurysms prior to rupture.
      MCA aneurysms were highlighted for their incidence rate and severe outcomes. The CASAM of morphological parameters is more precise and reliable than manual measurement to avoid bias.
      • Geng J.
      • Hu P.
      • Ji Z.
      • et al.
      Accuracy and reliability of computer-assisted semi-automated morphological analysis of intracranial aneurysms: an experimental study with digital phantoms and clinical aneurysm cases.
      We used CASAM to calculate the mirror MCA aneurysms precisely. We found that ruptured aneurysms had significantly greater D, WD, H, WH, BNF, SR, and AR. Hemodynamic analysis showed that minimum NWSS, mean NWSS and minimum WSSG were significantly smaller in the ruptured group. Ruptured aneurysms had larger maximum OSI, maximum CHP, mean CHP, maximum RRT, mean RRT, and LSA ratio. Size, N, and maximum OSI were independent factors. The AUC values of Size, N, and OSI were 0.794, 0.695, and 0.701, respectively.
      The results of our study involved more variables and were partially consistent with other studies.
      • van der Kamp L.T.
      • Rinkel G.J.E.
      • Verbaan D.
      • et al.
      Risk of rupture after intracranial aneurysm growth.
      • Elsharkawy A.
      • Lehečka M.
      • Niemelä M.
      • et al.
      Anatomic risk factors for middle cerebral artery aneurysm rupture: computed tomography angiography study of 1009 consecutive patients.
      • Qi P.
      • Feng X.
      • Lu J.
      • et al.
      Morphological irregularity of unruptured intracranial aneurysms is more related with aneurysm size rather than cerebrovascular atherosclerosis: a case-control study.
      The mirror aneurysm study of Xu et al. demonstrated that size, H, BNF, irregular shape, maximum WSS, and NWSS were associated with ruptured MCA aneurysm.
      • Xu L.
      • Wang H.
      • Chen Y.
      • et al.
      Morphological and hemodynamic factors associated with ruptured middle cerebral artery mirror aneurysms: a retrospective study.
      N, BNF, SR, and AR were also relevant factors in Xu et al.
      • Xu W.D.
      • Wang H.
      • Wu Q.
      • et al.
      Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
      Jiang et al. demonstrated the association of NWSS, LSA, WSSG, and OSI with rupture.
      • Jiang P.
      • Liu Q.
      • Wu J.
      • et al.
      A novel scoring system for rupture risk stratification of intracranial aneurysms: a hemodynamic and morphological study.
      A national-wide case-control study also showed the significance of the LSA ratio.
      • Skodvin T.
      • Evju Ø.
      • Helland C.A.
      • et al.
      Rupture prediction of intracranial aneurysms: a nationwide matched case-control study of hemodynamics at the time of diagnosis.
      In our study, maximum OSI value was the independent factor.

      Size

      Studies on mirror aneurysms were consistent regarding size and irregularity.
      • Han P.
      • Jin D.
      • Wei W.
      • et al.
      The prognostic effects of hemodynamic parameters on rupture of intracranial aneurysm: a systematic review and meta-analysis.
      ,
      • Maslehaty H.
      • Capone C.
      • Frantsev R.
      • et al.
      Predictive anatomical factors for rupture in middle cerebral artery mirror bifurcation aneurysms.
      Kamp et al. found that risk factors for aneurysm rupture during follow-up were size (> 7 mm) and site. The site in MCA had the highest hazard ratio in their study.
      • van der Kamp L.T.
      • Rinkel G.J.E.
      • Verbaan D.
      • et al.
      Risk of rupture after intracranial aneurysm growth.
      Elsharkawy et al. identified 7–14 mm as a risk factor.
      • Elsharkawy A.
      • Lehečka M.
      • Niemelä M.
      • et al.
      Anatomic risk factors for middle cerebral artery aneurysm rupture: computed tomography angiography study of 1009 consecutive patients.
      The ruptured mean size in our study was 8.16 mm, partially in line with their results. Even the irregular shape was related to the size. Qi et al. found that morphological irregularity was more closely correlated with size than cerebrovascular atherosclerosis in unruptured aneurysms.
      • Qi P.
      • Feng X.
      • Lu J.
      • et al.
      Morphological irregularity of unruptured intracranial aneurysms is more related with aneurysm size rather than cerebrovascular atherosclerosis: a case-control study.
      lower total volume ratio (TVR), then the blood in the aneurysm was not fully circulating, causing disturbed blood flow.
      • Lee U.Y.
      • Kwak H.S.
      Analysis of morphological-hemodynamic risk factors for aneurysm rupture including a newly introduced total volume ratio.
      We assume that the saccular aneurysm is a sphere. Size becomes a significant quantity in its calculation of volume. When size increases, TVR tends to decrease, causing hemodynamic changes in the lumen of IAs. The cutoff value of the Size was 6.30 mm, which help discriminate the rupture risk. Size determination is essential for treatment.

      Neck diameter

      The neck diameter (N) poses challenges for neurosurgeons.
      • Pierot L.
      • Spelle L.
      • Cognard C.
      • et al.
      Wide neck bifurcation aneurysms: what is the optimal endovascular treatment?.
      The relationship between N and rupture risk is debated.
      • Mocco J.
      • Brown R.D.
      • Torner J.C.
      • et al.
      Aneurysm morphology and prediction of rupture: an international study of unruptured intracranial aneurysms analysis.
      ,
      • Chen Y.
      • Xing H.
      • Lin B.
      • et al.
      Morphological risk model assessing anterior communicating artery aneurysm rupture: development and validation.
      Smaller N independently correlated with the rupture. A smaller N brought causes a more significant obstruction to the blood inside the aneurysm. Hypothetically, in mirror MCA aneurysms, each pair of the blood velocities and localized pressures would be equal, and a smaller N would cause complicated hemodynamics. Xu et al. found that smaller N increased shear stress.
      • Xu W.D.
      • Wang H.
      • Wu Q.
      • et al.
      Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
      Using CFD simulation, a higher velocity and more substantial impact were observed in the narrow-necked IAs.
      • Tang X.
      • Zhou L.
      • Wen L.
      • et al.
      Morphological and hemodynamic characteristics associated with the rupture of multiple intracranial aneurysms.
      Wide-necked aneurysms have been a challenge; Gory et al. showed that, after a 1-year follow-up, only 67.7 % (21/31) of wide-necked aneurysms were occluded entirely after treatment.
      • Gory B.
      • Aguilar-Pérez M.
      • Pomero E.
      • et al.
      One-year angiographic results after pCONus stent-assisted coiling of 40 wide-neck middle cerebral artery aneurysms.
      A wide neck increased the complexity of treatment and (to some extent) decreased the risk of rupture. In the same-sized mirror aneurysm cases, N was a critical variable. Future studies should assess ruptured aneurysms with different Ns and even unruptured aneurysms with different Ns.

      Maximum OSI

      Oscillatory shear index (OSI) is a time concept, which more comprehensively reflect the dynamic change.
      • Kim T.
      • Oh C.W.
      • Bang J.S.
      • et al.
      Higher oscillatory shear index is related to aneurysm recanalization after coil embolization in posterior communicating artery aneurysms.
      Kim et al. deemed that high OSI causes endothelial cell dysfunction which can affect the formation and the growth of the intracranial aneurysm.
      • Kim T.
      • Oh C.W.
      • Bang J.S.
      • et al.
      Higher oscillatory shear index is related to aneurysm recanalization after coil embolization in posterior communicating artery aneurysms.
      Although ruptured aneurysm had the unregular endothelial environment,
      • Frösen J.
      • Cebral J.
      • Robertson A.M.
      • et al.
      Flow-induced, inflammation-mediated arterial wall remodeling in the formation and progression of intracranial aneurysms.
      an elevated OSI was more related to the vascular remodeling.
      • Chen Z.
      • Yu H.
      • Shi Y.
      • et al.
      Vascular remodelling relates to an elevated oscillatory shear index and relative residence time in spontaneously hypertensive rats.
      Previous studies also emphasized that the association between the focal change of aneurysm with the mean OSI.
      • Cornelissen B.M.W.
      • Leemans E.L.
      • Slump C.H.
      • et al.
      Hemodynamic changes after intracranial aneurysm growth.
      ,
      • Cebral J.R.
      • Detmer F.
      • Chung B.J.
      • et al.
      Local hemodynamic conditions associated with focal changes in the intracranial aneurysm wall.
      In our study, mean OSI was insignificant, and maximum OSI as a protective role was independently correlated with the aneurysm rupture. Maximum OSI represented the highest oscillatory change in the aneurysm. It might enhance the focal wall, and could apply a protective basis for aneurysm rupture.

      Limitations

      This study has some limitations. First, retrospective studies are limited in ascertaining causality between features and outcomes. Inconsistent views have stayed on the change of the morphology after the rupture.
      • Rahman M.
      • Ogilvy C.S.
      • Zipfel G.J.
      • et al.
      Unruptured cerebral aneurysms do not shrink when they rupture: multicenter collaborative aneurysm study group.
      ,
      • Skodvin T.
      • Johnsen L.H.
      • Gjertsen Ø.
      • et al.
      Cerebral aneurysm morphology before and after rupture: nationwide case series of 29 aneurysms.
      Second, because of the rarity of mirror MCA aneurysms in a single institution, the limited sample size influenced the robustness of our findings. Even though we performed conditional logistic regression with the stepwise method, relative mall sample limited the results of the multivariate analysis. Third, even if studies showed CASAM had better reliability and precision, this measurement is not yet widely used. It may be challenging to achieve total agreement with studies of manual measurements. Fourth, there are no uniform hemodynamic parameter settings across institutions. We applied our hemodynamic thresholds and grid drawing methods to generate colorful images. This approach may cause significant parameters to appear less noticeable. Fifth, AI analysis may be better to evaluate the performance of risk factors.
      • Katsuki M.
      • Kakizawa Y.
      • Nishikawa A.
      • et al.
      Easily created prediction model using deep learning software (Prediction One, Sony Network Communications Inc.) for subarachnoid hemorrhage outcomes from small dataset at admission.
      Finally, the study of mirror aneurysms can neutralize the influence of aneurysm location. We only included mirror MCA aneurysms, which might not be generalizable to aneurysms in other locations. Further prospective studies are needed in several institutions.

      Conclusions

      Morphology and hemodynamics can help predict aneurysm rupture risks. The more significant size, smaller neck diameter and maximum OSI were independent risk factors for the rupture of MCA aneurysms. These variables could aid practical risk evaluation.

      Data Availability Statement

      The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

      Author contribution

      SH: Conceputualization, Methodology; writing; WX: Data curation, formal analysis, review&editing; RC: review&editing; JY: supervision. All authors contributed to the article and approved the submitted version.

      Publisher's Note

      All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

      Declaration of Competing Interest

      The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

      References

        • Neifert S.N.
        • Chapman E.K.
        • Martini M.L.
        • et al.
        Aneurysmal subarachnoid hemorrhage: the last decade.
        Transl Stroke Res. 2021; 12: 428-446
        • Xu W.D.
        • Chen R.D.
        • Hu S.Q.
        • et al.
        Morphological evaluation of the risk of posterior communicating artery aneurysm rupture: a mirror aneurysm model.
        J Neurosurg. 2022; : 1-6
        • Xu W.D.
        • Wang H.
        • Wu Q.
        • et al.
        Morphology parameters for rupture in middle cerebral artery mirror aneurysms.
        J Neurointerv Surg. 2020; 12: 858-861
        • Frösen J.
        • Cebral J.
        • Robertson A.M.
        • et al.
        Flow-induced, inflammation-mediated arterial wall remodeling in the formation and progression of intracranial aneurysms.
        Neurosurg Focus. 2019; 47: E21
        • Diagbouga M.R.
        • Morel S.
        • Bijlenga P.
        • et al.
        Role of hemodynamics in initiation/growth of intracranial aneurysms.
        Eur J Clin Invest. 2018; 48: e12992
        • Soldozy S.
        • Norat P.
        • Elsarrag M.
        • et al.
        The biophysical role of hemodynamics in the pathogenesis of cerebral aneurysm formation and rupture.
        Neurosurg Focus. 2019; 47: E11
        • Han P.
        • Jin D.
        • Wei W.
        • et al.
        The prognostic effects of hemodynamic parameters on rupture of intracranial aneurysm: a systematic review and meta-analysis.
        Int J Surg. 2021; 86: 15-23
        • Xu L.
        • Wang H.
        • Chen Y.
        • et al.
        Morphological and hemodynamic factors associated with ruptured middle cerebral artery mirror aneurysms: a retrospective study.
        World Neurosurg. 2020; 137: e138-e143
        • Maslehaty H.
        • Capone C.
        • Frantsev R.
        • et al.
        Predictive anatomical factors for rupture in middle cerebral artery mirror bifurcation aneurysms.
        J. Neurosurg. 2018; 128: 1799-1807
        • Hu S.Q.
        • Chen R.D.
        • Xu W.D.
        • et al.
        A predictive hemodynamic model based on risk factors for ruptured mirror aneurysms.
        Front Neurol. 2022; 13998557
        • Murayama Y.
        • Fujimura S.
        • Suzuki T.
        • et al.
        Computational fluid dynamics as a risk assessment tool for aneurysm rupture.
        Neurosurg Focus. 2019; 47: E12
        • Tang X.
        • Zhou L.
        • Wen L.
        • et al.
        Morphological and hemodynamic characteristics associated with the rupture of multiple intracranial aneurysms.
        Front Neurol. 2021; 12811281
        • Liu J.
        • Chen Y.
        • Zhu D.
        • et al.
        A nomogram to predict rupture risk of middle cerebral artery aneurysm.
        Neurol Sci. 2021; 42: 5289-5296
        • Geng J.
        • Hu P.
        • Ji Z.
        • et al.
        Accuracy and reliability of computer-assisted semi-automated morphological analysis of intracranial aneurysms: an experimental study with digital phantoms and clinical aneurysm cases.
        Int J Comput Assist Radiol Surg. 2020; 15: 1749-1759
        • Jeong W.
        • Han M.H.
        • Rhee K.
        The hemodynamic alterations induced by the vascular angular deformation in stent-assisted coiling of bifurcation aneurysms.
        Comput Biol Med. 2014; 53: 1-8
        • Ford M.D.
        • Alperin N.
        • Lee S.H.
        • et al.
        Characterization of volumetric flow rate waveforms in the normal internal carotid and vertebral arteries.
        Physiol Meas. 2005; 26: 477-488
        • Yuan J.
        • Huang C.
        • Lai N.
        • et al.
        Hemodynamic and morphological analysis of mirror aneurysms prior to rupture.
        Neuropsychiatr Dis Treat. 2020; 16: 1339-1347
        • van der Kamp L.T.
        • Rinkel G.J.E.
        • Verbaan D.
        • et al.
        Risk of rupture after intracranial aneurysm growth.
        JAMA Neurol. 2021; 78: 1228-1235
        • Elsharkawy A.
        • Lehečka M.
        • Niemelä M.
        • et al.
        Anatomic risk factors for middle cerebral artery aneurysm rupture: computed tomography angiography study of 1009 consecutive patients.
        Neurosurgery. 2013; 73 (discussion 836-827): 825-837
        • Qi P.
        • Feng X.
        • Lu J.
        • et al.
        Morphological irregularity of unruptured intracranial aneurysms is more related with aneurysm size rather than cerebrovascular atherosclerosis: a case-control study.
        Clin Interv Aging. 2021; 16: 665-674
        • Jiang P.
        • Liu Q.
        • Wu J.
        • et al.
        A novel scoring system for rupture risk stratification of intracranial aneurysms: a hemodynamic and morphological study.
        Front Neurosci. 2018; 12: 596
        • Skodvin T.
        • Evju Ø.
        • Helland C.A.
        • et al.
        Rupture prediction of intracranial aneurysms: a nationwide matched case-control study of hemodynamics at the time of diagnosis.
        J Neurosurg. 2018; 129: 854-860
        • Lee U.Y.
        • Kwak H.S.
        Analysis of morphological-hemodynamic risk factors for aneurysm rupture including a newly introduced total volume ratio.
        J Pers Med. 2021; 11
        • Pierot L.
        • Spelle L.
        • Cognard C.
        • et al.
        Wide neck bifurcation aneurysms: what is the optimal endovascular treatment?.
        J Neurointerv Surg. 2021; 13: e9
        • Mocco J.
        • Brown R.D.
        • Torner J.C.
        • et al.
        Aneurysm morphology and prediction of rupture: an international study of unruptured intracranial aneurysms analysis.
        Neurosurgery. 2018; 82 (Jr): 491-496
        • Chen Y.
        • Xing H.
        • Lin B.
        • et al.
        Morphological risk model assessing anterior communicating artery aneurysm rupture: development and validation.
        Clin Neurol Neurosurg. 2020; 197106158
        • Gory B.
        • Aguilar-Pérez M.
        • Pomero E.
        • et al.
        One-year angiographic results after pCONus stent-assisted coiling of 40 wide-neck middle cerebral artery aneurysms.
        Neurosurgery. 2017; 80: 925-933
        • Kim T.
        • Oh C.W.
        • Bang J.S.
        • et al.
        Higher oscillatory shear index is related to aneurysm recanalization after coil embolization in posterior communicating artery aneurysms.
        Acta Neurochir (Wien). 2021; 163: 2327-2337
        • Chen Z.
        • Yu H.
        • Shi Y.
        • et al.
        Vascular remodelling relates to an elevated oscillatory shear index and relative residence time in spontaneously hypertensive rats.
        Sci Rep. 2017; 7: 2007
        • Cornelissen B.M.W.
        • Leemans E.L.
        • Slump C.H.
        • et al.
        Hemodynamic changes after intracranial aneurysm growth.
        J Neurosurg. 2021; : 1-7
        • Cebral J.R.
        • Detmer F.
        • Chung B.J.
        • et al.
        Local hemodynamic conditions associated with focal changes in the intracranial aneurysm wall.
        AJNR Am J Neuroradiol. 2019; 40: 510-516
        • Rahman M.
        • Ogilvy C.S.
        • Zipfel G.J.
        • et al.
        Unruptured cerebral aneurysms do not shrink when they rupture: multicenter collaborative aneurysm study group.
        Neurosurgery. 2011; 68 (discussion 160-151): 155-160
        • Skodvin T.
        • Johnsen L.H.
        • Gjertsen Ø.
        • et al.
        Cerebral aneurysm morphology before and after rupture: nationwide case series of 29 aneurysms.
        Stroke. 2017; 48: 880-886
        • Katsuki M.
        • Kakizawa Y.
        • Nishikawa A.
        • et al.
        Easily created prediction model using deep learning software (Prediction One, Sony Network Communications Inc.) for subarachnoid hemorrhage outcomes from small dataset at admission.
        Surg Neurol Int. 2020; 11: 374