Performance of Radiomics derived morphological features for prediction of aneurysm rupture status

BackgroundMorphological differences between ruptured and unruptured cerebral aneurysms represent a focus of neuroimaging researchfor understanding the mechanisms of aneurysmal rupture. We evaluated the performance of Radiomics derived morphological features, recently proposed for rupture status classification, against automatically measured shape and size features previously established in the literature.Methods353 aneurysms (123 ruptured) from three-dimensional rotational catheter angiography (3DRA) datasets were analyzed. Based on a literature review, 13 Radiomics and 13 established morphological descriptors were automatically extracted per aneurysm, and evaluated for rupture status prediction using univariate and multivariate statistical analysis, yielding an area under the curve (AUC) metric of the receiver operating characteristic.ResultsValidation of overlapping descriptors for size/volume using both methods were highly correlated (p<0.0001, R2=0.99). Univariate analysis selected AspectRatio (p<0.0001, AUC=0.75), Non-sphericity Index (p<0.0001, AUC=0.75), Height/Width (p<0.0001, AUC=0.73), and SizeRatio (p<0.0001, AUC=0.73) as best among established descriptors, and Elongation (p<0.0001, AUC=0.71) and Flatness (p<0.0001, AUC=0.72) among Radiomics features. Radiomics Elongation correlated best with established Height/Width (R2=0.52), whereas Radiomics Flatness correlated best with Ellipticity Index (R2=0.54). Radiomics Sphericity correlated best with Undulation Index (R2=0.65). Best Radiomics performers, Elongation and Flatness, were highly correlated descriptors (p<0.0001, R2=0.75). In multivariate analysis, established descriptors (Height/Width, SizeRatio, Ellipticity Index; AUC=0.79) outperformed Radiomics features (Elongation, Maximum3Ddiameter; AUC=0.75).ConclusionAlthough recently introduced Radiomics analysis for aneurysm shape and size evaluation has the advantage of being an efficient operator independent methodology, it currently offers inferior rupture status discriminant performance compared with established descriptors. Future research is needed to extend the current Radiomics feature set to better capture aneurysm shape information.

Discrepancy between MRA and DSA in identifying the shape of small intracranial aneurysms

OBJECTIVEThe authors evaluated the sensitivity and accuracy of MRA in identifying the shape of small-sized unruptured intracranial aneurysms.METHODSSmall (< 7 mm) unruptured intracranial aneurysms initially detected by MRA and confirmed by DSA between January 2017 and December 2018 were morphologically reviewed by neuroradiologists. Regularity or irregularity of aneurysm shape was analyzed by two independent reviewers using MRA without DSA results. DSA findings served as the reference standard for aneurysm shape. Irregular shape, which in small aneurysms is associated with a higher likelihood of rupture, was defined as positive, and MRA sensitivity, specificity, and accuracy were determined by using evaluations based on location, size, and MRA magnetic strength (1.5T vs 3T MRA). Multivariate analysis was performed to determine risk factors for false-negative MRA results for irregularly shaped aneurysms.RESULTSIn total, 652 unruptured intracranial aneurysms in 530 patients were reviewed for this study. For detecting aneurysm shape irregularity, the overall MRA sensitivity was 60.4% for reviewer 1 and 60.9% for reviewer 2. Anterior cerebral artery aneurysms had the lowest sensitivity for location (36.7% for reviewer 1, 46.9% for reviewer 2); aneurysms sized < 3 mm had the lowest sensitivity for size (26.7% for both reviewers); and 1.5T MRA had lower sensitivity and accuracy than 3T MRA. In multivariate analysis, location, size, and magnetic strength of MRA were independent risk factors for false-negative MRA results for irregularly shaped aneurysms.CONCLUSIONSMRA had a low sensitivity for detecting the irregular shape of small intracranial aneurysms. In particular, anterior cerebral artery location, aneurysm size < 3 mm, and detection with 1.5T MRA were associated with a higher risk of irregularly shaped aneurysms being misjudged as regular.

Impact of aneurysm shape and neck configuration on cerebral infarction during microsurgical clipping of intracranial aneurysms

OBJECTIVECerebral infarction is a significant cause of morbidity and mortality related to microsurgical clipping of intracranial aneurysms. The objective of this study was to determine the impact of aneurysm shape and neck configuration on cerebral infarction after aneurysm surgery.METHODSThe authors retrospectively reviewed consecutive cases of ruptured and unruptured aneurysms treated with microsurgical clipping at their institution between 2010 and 2018. Three-dimensional reconstructions from preoperative computed tomography and digital subtraction angiography were used to determine aneurysm shape (regular/complex) and neck configuration (regular/irregular). Morphological and procedure-related risk factors for cerebral infarction were identified using univariate and multivariate statistical analyses.RESULTSAmong 243 patients with 252 aneurysms (148 ruptured, 104 unruptured), the overall cerebral infarction rate was 17.1%. Infarction tended to occur more often in aneurysms with complex shape (p = 0.084). Likewise, aneurysms with an irregular neck had a significantly higher rate of infarction (37.5%) than aneurysms with regular neck configuration (10.1%, p < 0.001). Aneurysms with an irregular neck were associated with a higher rate of intraoperative rupture (p = 0.003) and temporary parent artery occlusion (p = 0.037). In the multivariate analysis, irregular neck configuration was identified as an independent risk factor for infarction (OR 4.2, 95% CI 1.9–9.4, p < 0.001), whereas the association between aneurysm shape and infarction was not significant (p = 0.966).CONCLUSIONSIrregular aneurysm neck configuration represents an independent risk factor for cerebral infarction during microsurgical clipping of both ruptured and unruptured aneurysms.

Predictive anatomical factors for rupture in middle cerebral artery mirror bifurcation aneurysms

OBJECTIVEThe aim of this study was to define predictive factors for rupture of middle cerebral artery (MCA) mirror bifurcation aneurysms.METHODSThe authors retrospectively analyzed the data in patients with ruptured MCA bifurcation aneurysms with simultaneous presence of an unruptured MCA bifurcation mirror aneurysm treated in two neurosurgical centers. The following parameters were measured and analyzed with the statistical software R: neck, dome, and width of both MCA aneurysms—including neck/dome and width/neck ratios, shape of the aneurysms (regular vs irregular), inflow angle of both MCA aneurysms, and the diameters of the bilateral A1 and M1 segments and the frontal and temporal M2 trunks, as well as the bilateral diameter of the internal carotid artery (ICA).RESULTSThe authors analyzed the data of 44 patients (15 male and 29 female, mean age 50.1 years). Starting from the usual significance level of 0.05, the Sidak-corrected significance level is 0.0039. The diameter of the measured vessels was statistically not significant, nor was the inflow angle. The size of the dome was highly significant (p = 0.0000069). The size of the neck (p = 0.0047940) and the width of the aneurysms (p = 0.0056902) were slightly nonsignificant at the stated significance level of 0.0039. The shape of the aneurysms was bilaterally identical in 22 cases (50%). In cases of asymmetrical presentation of the aneurysm shape, 19 (86.4%) ruptured aneurysms were irregular and 3 (13.6%) had a regular shape (p = 0.001).CONCLUSIONSIn this study the authors show that the extraaneurysmal flow dynamics in mirror aneurysms are nonsignificant, and the aneurysmal geometry also does not seem to play a role as a predictor for rupture. The only predictors for rupture were size and shape of the aneurysms. It seems as though under the same conditions, one of the two aneurysms suffers changes in its wall and starts growing in a more or less stochastic manner. Newer imaging methods should enable practitioners to see which aneurysm has an unstable wall, to predict the rupture risk. At the moment one can only conclude that in cases of MCA mirror aneurysms the larger one, with or without shape irregularities, is the unstable aneurysm and that this is the one that needs to be treated.

Critical role of angiographic acquisition modality and reconstruction on morphometric and haemodynamic analysis of intracranial aneurysms

BackgroundSubtracted 3-D rotational angiography (3DRA) and cone-beam computed tomography angiography (CBCT-A) are often used in assessing cerebral aneurysm shape and haemodynamic profile. We sought to evaluate the effect of imaging modality, reconstruction parameters, and kernel selection on patient-derived aneurysm morphology and computational fluid dynamic (CFD) analysis to assess its potential contribution to inter-study variability.MethodsFour patients (five aneurysms) underwent concurrent 3DRA followed by high-resolution CBCT-A. Six models were reconstructed per aneurysm: 3DRA reconstructed with 0.28 and 0.14 mm voxel sizes (large and small volume of interest (VOI) respectively), and two kernel types (normal/smooth). CBCT-A was reconstructed over small VOI using normal/sharp kernel. Maximal dome dimension, neck diameter and dome/neck ratio were evaluated in 3D. Wall shear stress (WSS) magnitude was evaluated on the entire aneurysm dome and in the 5% dome areas covered by lowest (LWSS) and highest (HWSS) WSS. Parameters were evaluated with pairwise t-test analysis.ResultsSmaller VOI reconstructions resulted in smaller Dmax (P value=0.03) and Dmax/neck (P value=0.006) and in larger LWSS (P value=0.03). Use of sharp kernel led to narrower neck (P value=0.04) and higher Dmax/neck values (P value=0.02). CBCT-A resulted in statistically different aneurysm shape (up to 24% difference) and haemodynamics (up to 97% difference) compared with 3DRA.ConclusionThe choice of catheter 3D angiographic modality and reconstruction kernel has a critical impact on derived aneurysm morphological and haemodynamic analysis. The resultant variability can confound and obscure underlying differences within patient populations and between studies performed at different centres using divergent techniques, compromising the accuracy of quantitative aneurysm analysis.

Determinants of Gadolinium-Enhancement of the Aneurysm Wall in Unruptured Intracranial Aneurysms

BACKGROUND Recent studies suggest that gadolinium-enhancement of the aneurysm wall may reflect aneurysm wall inflammation, which might increase the risk of aneurysm growth or rupture. OBJECTIVE To determine the prevalence of gadolinium-enhancement and its determinants in patients with predominantly small (&lt;7 mm) unruptured intracranial aneurysms. METHODS We performed 3 T MRI aneurysm wall imaging in patients with ≥1 unruptured aneurysm. With Poisson regression analysis, we calculated crude and adjusted risk ratios (RRs) with 95% confidence intervals (CI) to assess determinants of gadolinium-enhancement. RESULTS Gadolinium-enhancement was observed in 25/79 patients (32%; 95% CI: 21%-42%) and 26/89 aneurysms (29%; 95% CI: 20%-39%). With aneurysms 1.0-2.9 mm as reference, RRs were 4.6 (95% CI: 0.6-36.5) for 3.0 to 4.9 mm, 9.4 (95% CI: 1.3-67.2) for 5.0 to 6.9 mm, and 14.8 (95% CI: 2.1-104.6) for ≥7.0 mm. With internal carotid artery as reference, RRs adjusted for aneurysm size were 3.6 (95% CI: 1.1-11.4) for posterior communicating artery and 3.0 (95% CI: 1.0-8.6) for middle cerebral artery. RRs were 0.8 (95% CI: 0.4-1.8) for acetylsalicylic acid use, 0.9 (95% CI: 0.5-1.8) for statin use, 1.4 (95% CI: 0.7-2.8) for hypertension, 0.9 (95% CI: 0.4-1.8) for previous smoking, 0.9 (95% CI: 0.3-2.6) for never smoking, and 1.4 (95% CI: 0.7-2.7) for irregular aneurysm shape. CONCLUSION Approximately one-third of patients had gadolinium-enhancement. Because aneurysm size is the strongest determinant of gadolinium-enhancement and also an established predictor for aneurysm growth and rupture, prospective studies with serial imaging need to investigate if gadolinium-enhancement predicts aneurysm growth and rupture.

Optimal first coil selection to avoid aneurysmal recanalization in endovascular intracranial aneurysmal coiling

BackgroundAneurysmal recanalization is a problem with endovascular coiling and one of its risk factors is the low volume embolization ratio (VER). The first coil VER (1st VER) is believed to be critical for obtaining a high VER. The main objective of this study was to evaluate factors potentially useful for selecting the optimal 1st VER for endovascular coiling.Methods609 initial saccular aneurysmal treatments performed between January 2010 and December 2014 at our institution were included in this retrospective study. Attempted procedures, retreatment cases, intraoperative rupture cases, and stent-assisted coiling cases were excluded. Age, sex, aneurysm location, ruptured aneurysm, aneurysm shape, neck size, maximum aneurysm size, dome-to-neck ratio, aneurysm volume, procedure, immediate Raymond scale score, 1st VER, and VER between the recanalization groups and non-recanalization groups were compared.ResultsThe factors related to recanalization were ruptured aneurysms, neck width, maximum aneurysm size, aneurysm volume, procedure, 1st VER, and VER. The cut-off values for aneurysm recanalization were a 1st VER of 10.0% and a VER of 33.0%. The maximum average VER of normal size aneurysms was found in the groups with a 1st VER of 17.5–20.0%.Conclusions1st VER was found to be a helpful index for estimating aneurysmal recanalization after coil embolization. The target 1st VER was 17.5–20.0% for obtaining a higher VER and avoiding recanalization.

Assessment of intracranial aneurysm rupture based on morphology parameters and anatomical locations

ObjectivesThe aim of the present study was to identify image-based morphological parameters and anatomical locations associated with intracranial aneurysm (IA) rupture.MethodsNine morphological parameters and aneurysm location were evaluated in 150 patients with saccular IAs (82 unruptured, 68 ruptured) using three-dimensional geometry. Aneurysm location and morphological parameters including size, aspect ratio, size ratio, height–width ratio, flow angle, aneurysm inclination angle, parent artery angle, vessel angle, and aneurysm shape were explored to identify a correlation with aneurysm rupture. These factors were analyzed using a two-tailed independent Student t test or the χ test for significance. Significant factors were further examined using logistic regression analysis. Additionally, receiver operating characteristic (ROC) analysis was performed to evaluate each parameter.ResultsStatistically significant differences were observed in ruptured and unruptured groups for aspect ratio, size ratio, height–width ratio, flow angle, aneurysm inclination angle, vessel angle, aneurysm shape, and aneurysm location. Logistic regression analysis further revealed that size ratio (OR 1.66; 95% CI 1.05 to 2.64), height–width ratio (OR 14.22; 95% CI 2.67 to 75.88), aneurysm inclination angle (OR 1.04; 95% CI 1.01 to 1.07), aneurysm shape (OR 4.68; 95% CI 2.44 to 8.98), and aneurysm location (OR 1.60; 95% CI 1.15 to 2.23) had the strongest independent correlation with ruptured IA. The ROC analysis showed that the size ratio and flow angle had the highest area under the curve, with values of 0.735 and 0.730, respectively.ConclusionsSize ratio, height–width ratio, aneurysm inclination angle, aneurysm shape, and aneurysm location might be important for discriminating between ruptured and unruptured aneurysms. Further investigation will determine whether these morphological parameters and anatomical locations will be reliable predictors of aneurysm rupture.