Clinicians managing unruptured cerebrovascular lesions regularly face difficult clinical choices. Accurately determining intracranial aneurysm rupture risk represents one of the most critical challenges in contemporary neurovascular practice. While preventive surgical clipping or endovascular embolization prevents subarachnoid hemorrhage, both procedural options carry inherent therapeutic risks. Therefore, objective noninvasive stratification tools are essential for guiding therapeutic interventions. In routine settings, clinicians often rely on conventional dimensional thresholds, such as maximum dome diameter. However, size alone frequently fails to capture the intricate hemodynamic and structural alterations preceding aneurysm rupture. To address this clinical limitation, modern neuroimaging protocols increasingly explore advanced computational parameters. A recent investigation in Clinical Neurology and Neurosurgery evaluated automated morphological features and threshold-based computed tomography perfusion metrics derived from one-stop computed tomography angiography and computed tomography perfusion. By analyzing these multi-parametric inputs, researchers sought to establish whether specific geometric or perfusion phenotypes correlate reliably with clinical outcomes.

Automated Imaging in Assessing Intracranial Aneurysm Rupture Risk

Computed tomography angiography protocols have advanced significantly with automated post-processing platforms. These automated platforms rapidly extract complex geometric descriptors without operator bias. Historically, clinicians manually measured simple variables such as maximum dome height and aspect ratio. However, manual evaluations introduce notable inter-observer variability that can obscure subtle morphological changes. Automated volumetric analysis now enables objective quantification of three-dimensional vascular architecture. In the recent retrospective study, researchers evaluated 60 patients with large unruptured intracranial aneurysms who underwent one-stop imaging between 2013 and 2023. Over a structured three-year follow-up period, eight patients experienced spontaneous aneurysm rupture. To address statistical challenges inherent in rare-event cohorts, investigators employed Firth penalized logistic regression. This mathematical approach effectively reduces small-sample bias during parameter estimation. Furthermore, automated algorithms segmented vascular boundaries and dome topography without manual intervention. Consequently, the team extracted robust morphological metrics alongside regional microvascular perfusion datasets. This noninvasive workflow represents an important methodological transition from qualitative visual assessment to objective computational phenotyping. Therefore, automated feature extraction provides a standardized foundation for future clinical risk calculators.

Contrasting Morphologic Metrics and Computed Tomography Perfusion

The investigators conducted a head-to-head comparison between anatomical shape indices and threshold-based perfusion parameters. Clinicians have long hypothesized that perianeurysmal parenchymal hypoperfusion or altered local microvascular dynamics might reflect wall instability. However, the study findings demonstrated that perfusion metrics exhibited no statistically significant association with rupture events. In contrast, computed tomography angiography morphologic parameters demonstrated robust discriminative signals. Specifically, posterior circulation location, undulation index, and nonsphericity index correlated positively with heightened rupture status. Meanwhile, neck diameter and volume-to-neck area ratio exhibited an inverse association with aneurysm rupture. The failure of perfusion metrics to provide diagnostic value highlights important physiological nuances. Computed tomography perfusion thresholds capture capillary transit times and regional parenchymal blood volume. Yet, aneurysm wall degradation occurs primarily through localized endothelial dysfunction, transmural inflammation, and mechanical shear stress. Because parenchymal perfusion reflects distal microcirculatory resistance rather than intramural tensile failure, its signals do not directly capture wall vulnerability. Consequently, structural luminal geometry remains far more informative than brain perfusion indices when evaluating unruptured lesions.

Diagnostic Performance of Undulation Index and Neck Diameter

To develop a clinically actionable screening tool, the researchers investigated bivariable combinations of top-performing variables. Among ten distinct bivariable configurations, the pairing of neck diameter and undulation index yielded exceptional diagnostic performance. The unadjusted receiver operating characteristic curve generated an area under the curve of 0.911. Furthermore, after rigorous bootstrap internal validation, the corrected area under the curve remained impressive at 0.888. The model also demonstrated an excellent bootstrap-corrected Brier score of 0.098, confirming high calibration accuracy. In this final predictive model, undulation index served as an independent positive risk factor, whereas neck diameter retained a strong protective association. From a geometric perspective, wide aneurysm necks reduce hemodynamic wall shear stress concentration across the lesion dome. Conversely, narrow necks promote complex intra-aneurysmal blood recirculation and higher inflow jet velocities. Consequently, lesions combining narrow necks with highly undulating, irregular sacs experience severe mechanical stresses. By combining these two automated variables, clinicians gain a parsimonious computational model that separates unstable lesions from indolent vascular outpouchings.

Biomechanical Implications of Geometric Aneurysm Irregularity

Aneurysm morphology directly reflects underlying wall biomechanics and chronic biological remodeling. The undulation index quantifies surface irregularities and lobular undulations across the aneurysm surface. Healthy arterial walls maintain uniform lamellar architecture under laminar flow conditions. However, abnormal geometric undulations induce turbulent blood recirculation patterns, stagnant flow zones, and erratic oscillatory shear stresses. In response to altered shear stresses, endothelial cells undergo apoptosis and downregulate protective nitric oxide synthase. Concurrently, inflammatory macrophages infiltrate the tunica media and secrete matrix metalloproteinases. These proteolytic enzymes actively degrade collagen bundles and elastin fibers, leading to focal wall thinning and secondary bleb formation. Because the undulation index numerically captures these localized protuberances, it acts as an indirect surrogate for progressive tissue degradation. Similarly, a high nonsphericity index identifies asymmetrical dome expansion driven by directional hemodynamic impact. Therefore, geometric irregularities detected on computed tomography angiography indicate advanced tissue remodeling. Identifying these biomechanical markers allows neurovascular teams to detect structurally fragile aneurysms before catastrophic hemorrhage occurs.

Translating Exploratory Computational Findings into Clinical Practice

Although the bivariable model demonstrated exceptional discrimination, neurosurgeons and neurologists must interpret these preliminary findings prudently. The analyzed cohort contained only 60 patients and eight verified rupture events. Because small retrospective datasets carry risk of overfitting, these findings remain hypothesis-generating rather than clinically deployable tools. Clinicians cannot yet replace established clinical risk scoring systems, such as PHASES and UIATS, with unvalidated algorithms. Currently, established guidelines prioritize maximum size, patient age, hypertension history, and lesion location during intervention decisions. Nevertheless, incorporating automated geometric metrics could refine risk stratification in borderline clinical scenarios. For example, intermediate-sized aneurysms measuring five to seven millimeters often pose treatment dilemmas for surgical teams. If future prospective trials validate the undulation index, automated morphologic profiling could distinguish aggressive lesions needing immediate coiling from benign lesions suitable for conservative surveillance. Ultimately, multidisciplinary teams in neurovascular clinics will benefit from combining automated artificial intelligence segmentation with clinical assessment. Collaborative multicenter studies must now evaluate these geometric markers across diverse patient cohorts.

Frequently Asked Questions

How does undulation index indicate intracranial aneurysm rupture risk?

Undulation index measures the surface irregularity and geometric deviation of an aneurysm dome. Higher values reflect uneven sac morphology, localized wall thinning, and disturbed intra-aneurysmal hemodynamic shear stresses. Consequently, elevated undulation indicates structurally unstable tissue regions prone to progressive remodeling and spontaneous mechanical failure.

Why did computed tomography perfusion metrics show limited predictive value?

Threshold-based computed tomography perfusion metrics primarily capture microvascular parenchymal flow rather than focal mechanical wall stress. Furthermore, systemic hemodynamic fluctuations and volume averaging artifact within vascular lumina limit threshold accuracy. Therefore, macroscopic structural geometry reflects intrinsic lesion wall vulnerability far better than parenchymal perfusion signals in unruptured aneurysms.

Can clinicians use this bivariable model immediately for treatment decisions?

Clinicians cannot immediately apply this bivariable model to direct clinical decision-making. Although the internal bootstrap discrimination was high, the exploratory study evaluated a relatively small cohort with few rupture events. Thus, clinicians must await prospective external validation across diverse patient populations before adopting these automated algorithms.

Disclaimer: This content is for informational and educational purposes only and does not constitute medical advice. Consult qualified healthcare professionals for diagnosis and treatment. Refer to the latest local and national guidelines for clinical practice.

References

1. Niu J et al. Exploratory comparative analysis of automated morphologic and perfusion features derived from one-stop computed tomography angiography and computed tomography perfusion in large unruptured intracranial aneurysms. Clin Neurol Neurosurg. 2026 Oct. doi: 10.1016/j.clineuro.2026.109547. PMID: 42322734.
2. Dhar S, Tremmel M, Mocco J, et al. Morphology parameters for intracranial aneurysm rupture risk assessment. Neurosurgery. 2008;63(2):185-197.
3. Etminan N, Rinkel GJ. Unruptured intracranial aneurysms: development, rupture and prevention. Nat Rev Neurol. 2016;12(12):699-713.
4. Greving JP, Wermer MJ, Brown RD Jr, et al. Development of the PHASES score for prediction of risk of rupture of intracranial aneurysms: a pooled analysis of six prospective cohort studies. Lancet Neurol. 2014;13(1):59-66.