arxiv: 2604.06649 · v1 · submitted 2026-04-08 · ⚛️ physics.med-ph

Recognition: 2 theorem links

· Lean Theorem

Bayesian Aneurysm Growth Detection via Surface Displacement Modeling

Jorge A. Roa Castro , Abhishek Singh , Atharva Hans , Kostiantyn Kondratiuk , David Saloner , Vitaliy L. Rayz , Pavlos P. Vlachos , Ilias Bilionis

Authors on Pith no claims yet

Pith reviewed 2026-05-10 18:15 UTC · model grok-4.3

classification ⚛️ physics.med-ph

keywords intracranial aneurysmsgrowth detectionBayesian modelingsurface displacementMRAlongitudinal imagingprobabilistic assessment

0 comments

The pith

A Bayesian model using normal-directed displacements and the surrounding vessel as reference detects aneurysm growth with AUC 0.86-0.87 and improves expert agreement over volume measures.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a probabilistic method to identify growth in unruptured intracranial aneurysms from serial MRA scans. It registers baseline and follow-up vessel surfaces, measures normal-directed displacements at each point, and subtracts the average displacement on the nearby non-aneurysmal vessel from the average on the aneurysm itself. This difference serves as the growth signal inside a Bayesian framework that returns calibrated probabilities and uncertainty ranges. Evaluated on two longitudinal cohorts, the approach reaches AUC values of 0.86-0.87 and raises Cohen's kappa from 0.35 (volumetric rules) to 0.66, while remaining stable even when trained on labels from less experienced raters. The result is an interpretable, scanner-robust way to quantify subtle three-dimensional change without depending on manual diameter calls.

Core claim

We show that a Bayesian displacement-based model using the surrounding vessel as an internal reference achieves strong discrimination of aneurysm growth (AUC 0.86-0.87) and improves agreement with expert labels (Cohen's kappa up to 0.66 vs. 0.35 for volumetric criteria), while providing calibrated posterior probabilities with uncertainty bounds. The method registers baseline and follow-up surfaces, computes normal-directed displacements, and summarizes change as the difference between mean aneurysm displacement and mean displacement on the surrounding non-aneurysmal vessel segment.

What carries the argument

Bayesian model of normal-directed surface displacements that subtracts mean aneurysm displacement from mean displacement on the adjacent vessel segment to isolate true growth from imaging and registration artifacts.

If this is right

Calibrated posterior probabilities allow clinicians to set risk-adjusted thresholds for intervention or repeat imaging in borderline cases.
Robustness to lower-expertise labels supports deployment across centers with varying rater experience.
Probabilistic outputs with uncertainty bounds reduce reliance on binary volume or diameter rules that miss subtle surface change.
Cross-sequence and cross-scanner performance enables consistent surveillance protocols without retraining per site.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same displacement-difference approach could be tested on other longitudinal vascular datasets, such as aortic or carotid aneurysms, to check whether the internal-reference principle generalizes.
High-uncertainty cases flagged by the model could be routed automatically for additional imaging sequences or higher-resolution acquisitions.
Incorporating the posterior growth probability into existing rupture-risk calculators might improve patient-specific decision thresholds without requiring new clinical trials for every scanner type.

Load-bearing premise

The surrounding non-aneurysmal vessel segment experiences no meaningful structural change between the two scans, so its average displacement can cancel out shared imaging and processing errors.

What would settle it

A new cohort in which independent high-resolution imaging shows clear vessel-wall remodeling or movement in the reference segment and the model's AUC for growth detection falls below 0.75.

Figures

Figures reproduced from arXiv: 2604.06649 by Abhishek Singh, Atharva Hans, David Saloner, Ilias Bilionis, Jorge A. Roa Castro, Kostiantyn Kondratiuk, Pavlos P. Vlachos, Vitaliy L. Rayz.

**Figure 2.** Figure 2: Posterior thresholds and within-cohort predictions across raters. Rows correspond to the training reference: (a) external rater (MNHHS), (b) junior rater (UCSF), (c) senior rater (UCSF). Left: posterior density of the soft-threshold parameter τ expressed in millimetres; dashed line marks the posterior median. Middle: predicted growth probability P(growth) versus mean-shift di (mm) for all cases in the sam… view at source ↗

**Figure 3.** Figure 3: Models performance against labels from all three raters. AUC and Cohen’s [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗

**Figure 4.** Figure 4: Representative growing internal carotid artery (ICA) bifurcation aneurysm. [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗

**Figure 5.** Figure 5: Representative UCSF cases. Left: (top) baseline 2D cross-section with manual [PITH_FULL_IMAGE:figures/full_fig_p018_5.png] view at source ↗

**Figure 6.** Figure 6: Posterior soft-threshold across sites and scales. Posteriors of the cut-off parameter τ for models trained on external rater (MNHHS, time-of-flight MRA), senior rater (UCSF, contrast-enhanced MRA), and physician (UCSF, contrast-enhanced MRA) reference labels. Left: τ expressed in millimetres. Right: the same posteriors on the standardized scale z = (d − µ)/σ. Dashed lines indicate posterior medians. This f… view at source ↗

read the original abstract

Clinical decisions for unruptured intracranial aneurysms depend on detecting growth on follow-up magnetic resonance angiography (MRA). Growth is typically judged from manual 2D diameters on few slices, which vary across clinicians and frequently miss subtle 3D change. Even with 3D segmentations, apparent differences can reflect resolution, segmentation, surface processing, or registration mismatch rather than true growth; most criteria remain heuristic and binary. We show that a Bayesian displacement-based model using the surrounding vessel as an internal reference achieves strong discrimination of aneurysm growth (AUC 0.86-0.87) and improves agreement with expert labels (Cohen's kappa up to 0.66 vs. 0.35 for volumetric criteria), while providing calibrated posterior probabilities with uncertainty bounds. The method registers baseline and follow-up surfaces, computes normal-directed displacements, and summarizes change as the difference between mean aneurysm displacement and mean displacement on the surrounding non-aneurysmal vessel segment. The vessel segment serves as an internal control for imaging and processing variability, assuming negligible structural change over the surveillance interval. We evaluate two cohorts spanning time-of-flight and contrast-enhanced longitudinal MRA studies: a public dataset labeled from neuroradiologist-provided measurements and an institutional dataset labeled by senior and junior raters. Performance is preserved when training on lower-expertise labels, indicating robustness to label variability. Calibrated probabilities may aid clinical decision-making in borderline cases, where high uncertainty can motivate repeat imaging. This framework provides interpretable probabilistic growth assessment from longitudinal MRA, reduces dependence on clinician expertise, and supports cross-center surveillance across scanners and angiography sequences.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Bayesian vessel-referenced displacement model improves expert agreement on aneurysm growth but the reference assumption lacks direct checks.

read the letter

The main thing here is a Bayesian model that registers baseline and follow-up aneurysm surfaces from MRA, computes normal-directed displacements, and subtracts the mean displacement on the surrounding non-aneurysmal vessel to isolate growth. This produces AUCs of 0.86-0.87 and Cohen's kappa up to 0.66, beating the 0.35 from volumetric criteria, plus calibrated posterior probabilities with uncertainty bounds. They test on both a public dataset and an institutional one with senior and junior labels, and performance holds when using the lower-expertise labels. The vessel reference is the practical novelty for handling imaging and processing variability without external controls. The probabilistic framing is also useful for borderline cases where uncertainty might prompt repeat scans. The modeling steps read as grounded, with no circular definitions in the growth metric and appropriate citations for the imaging and Bayesian pieces. The soft spot is the vessel-as-control assumption. The method treats any vessel displacement as noise to be subtracted, which is fine only if the vessel segment shows negligible biological or imaging change over the interval. The abstract states this but the evaluation uses only expert growth labels; there are no reported vessel-only displacement stats, time-interval correlations, or sensitivity runs with simulated vessel drift. That leaves the performance numbers a bit harder to interpret if the assumption does not always hold. This is aimed at researchers in medical image analysis and clinicians doing aneurysm surveillance. Readers working on longitudinal vascular imaging or probabilistic decision support would get concrete value from the uncertainty outputs and the robustness checks. It deserves a serious referee because the clinical problem is high-stakes and the method offers a clear step past binary heuristics, even if the reference validation needs tightening. I would send it for peer review with a request to add direct tests on the vessel stability assumption.

Referee Report

1 major / 3 minor

Summary. The manuscript proposes a Bayesian displacement-based model for detecting growth in unruptured intracranial aneurysms from longitudinal MRA. Surfaces are registered between baseline and follow-up scans, normal-directed displacements are computed, and growth is quantified as the difference between mean aneurysm displacement and mean displacement on the adjacent non-aneurysmal vessel segment (treated as an internal control). The method is evaluated on a public dataset and an institutional cohort, reporting AUC 0.86-0.87 for growth discrimination, Cohen's kappa up to 0.66 (vs. 0.35 for volumetric criteria), and calibrated posterior probabilities with uncertainty bounds.

Significance. If the central results hold after addressing the load-bearing assumption, the work would provide a more objective, probabilistic alternative to manual diameter or volumetric criteria for aneurysm surveillance. The internal-reference construction and uncertainty quantification are strengths that could reduce inter-observer variability and support decisions in borderline cases across different scanners and sequences.

major comments (1)

[Abstract and Methods] Abstract and Methods (vessel reference construction): The growth metric is defined as the difference in mean normal-directed displacements between the aneurysm surface and the surrounding vessel segment. This construction is valid only under the assumption of negligible vessel change over the surveillance interval, yet no auxiliary analysis (vessel-only displacement statistics, correlation with interval length, or sensitivity to simulated vessel drift) is reported to bound possible violations. Because any systematic vessel displacement is subtracted from the aneurysm signal, this untested assumption is load-bearing for the reported AUC and kappa values.

minor comments (3)

[Results] Results: Report AUC and kappa values with confidence intervals or p-values for the improvement over volumetric baselines to allow assessment of statistical significance.
[Methods] Methods: Provide more detail on the surface registration algorithm, the exact definition of the vessel segment boundaries, and the Bayesian model hyperparameters to support reproducibility.
[Figures] Figure captions: Ensure all figures showing posterior distributions or displacement maps include scale bars, color legends, and explicit indication of which surfaces correspond to aneurysm versus vessel.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comment on the vessel reference construction. We address the concern directly below and will revise the manuscript to strengthen the supporting evidence for the central assumption.

read point-by-point responses

Referee: [Abstract and Methods] Abstract and Methods (vessel reference construction): The growth metric is defined as the difference in mean normal-directed displacements between the aneurysm surface and the surrounding vessel segment. This construction is valid only under the assumption of negligible vessel change over the surveillance interval, yet no auxiliary analysis (vessel-only displacement statistics, correlation with interval length, or sensitivity to simulated vessel drift) is reported to bound possible violations. Because any systematic vessel displacement is subtracted from the aneurysm signal, this untested assumption is load-bearing for the reported AUC and kappa values.

Authors: We agree that the assumption of negligible structural change in the non-aneurysmal vessel segment is load-bearing for the growth metric and the reported performance metrics. The original manuscript states the assumption explicitly but does not provide the auxiliary analyses suggested. In the revised manuscript we will add: (i) summary statistics of mean normal-directed displacements on the vessel-only segments for both cohorts, (ii) the correlation between vessel displacement and surveillance interval length, and (iii) a sensitivity analysis that injects controlled vessel drift and quantifies the resulting change in AUC and Cohen’s kappa. These additions will bound the possible impact of violations and increase confidence in the internal-reference approach. revision: yes

Circularity Check

0 steps flagged

No significant circularity; growth metric defined from physical reference assumption with external validation against expert labels

full rationale

The paper proposes a displacement-based growth metric by registering surfaces, computing normal-directed displacements, and taking the difference between mean aneurysm displacement and mean vessel-segment displacement. This construction rests on an explicit physical assumption (vessel stability as internal control) rather than defining the metric in terms of itself or fitting parameters to the target labels and then re-presenting the fit as a prediction. Evaluation uses independent expert labels on two separate cohorts to compute AUC and Cohen's kappa; no step reduces the reported performance to a tautology or to a self-citation chain. No uniqueness theorems, ansatzes smuggled via prior work, or renaming of known results are invoked. The Bayesian modeling step provides posterior probabilities and uncertainty but does not alter the external grounding of the performance claims.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach relies on the domain assumption of stable vessel geometry and the technical assumption that surface registration accurately captures true displacements without artifacts.

axioms (1)

domain assumption The surrounding non-aneurysmal vessel segment undergoes negligible structural change over the surveillance interval
This allows the vessel to serve as an internal control for imaging and processing variability.

pith-pipeline@v0.9.0 · 5625 in / 1293 out tokens · 76712 ms · 2026-05-10T18:15:11.302413+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean absolute_floor_iff_bare_distinguishability unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The vessel segment serves as an internal control for imaging and processing variability, assuming negligible structural change over the surveillance interval... d_i = a_i - v_i ... Bayesian soft-threshold model
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

normal-directed displacements... logistic(s(d*_i - τ))

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

34 extracted references · 28 canonical work pages

[1]

Inoue, H

T. Inoue, H. Shimizu, M. Fujimura, A. Saito, T. Tominaga, Annual rup- ture risk of growing unruptured cerebral aneurysms detected by mag- netic resonance angiography: Clinical article, Journal of Neurosurgery 117 (1) (2012) 20–25. doi:10.3171/2012.4.JNS112225

work page doi:10.3171/2012.4.jns112225 2012
[2]

W. A. Mehan, J. M. Romero, J. A. Hirsch, D. J. Sabbag, R. G. Gon- zalez, J. J. Heit, P. W. Schaefer, Unruptured intracranial aneurysms conservatively followed with serial CT angiography: could morphology and growth predict rupture?, Journal of NeuroInterventional Surgery 6 (10) (2014) 761–766. doi:10.1136/neurintsurg-2013-010944. 23

work page doi:10.1136/neurintsurg-2013-010944 2014
[3]

L. T. van der Kamp, G. J. E. Rinkel, D. Verbaan, R. van den Berg, W. P. Vandertop, Y. Murayama, T. Ishibashi, A. Lindgren, T. Koivisto, M. Teo, J. St George, R. Agid, I. Radovanovic, J. Moroi, K. Igase, I. R. van den Wijngaard, M. Rahi, J. Rinne, J. Kuhmonen, H. D. Boogaarts, G. K. C. Wong, J. M. Abrigo, A. Morita, Y. Shiokawa, K. A. M. Hacken- berg, N. E...

work page doi:10.1001/jamaneurol.2021.2915 2021
[4]

Brinjikji, Y.-Q

W. Brinjikji, Y.-Q. Zhu, G. Lanzino, H. Cloft, M. Murad, Z. Wang, D. Kallmes, Risk Factors for Growth of Intracranial Aneurysms: A Sys- tematic Review and Meta-Analysis, AJNR: American Journal of Neu- roradiology 37 (4) (2016) 615–620. doi:10.3174/ajnr.A4575

work page doi:10.3174/ajnr.a4575 2016
[5]

Etminan, G

N. Etminan, G. J. Rinkel, Unruptured intracranial aneurysms: develop- ment, rupture and preventive management, Nature Reviews Neurology 12 (12) (2016) 699–713. doi:10.1038/nrneurol.2016.150

work page doi:10.1038/nrneurol.2016.150 2016
[6]

A. M. Algra, A. Lindgren, M. D. I. Vergouwen, J. P. Greving, I. C. van der Schaaf, T. P. C. van Doormaal, G. J. E. Rinkel, Procedural Clin- ical Complications, Case-Fatality Risks, and Risk Factors in Endovascu- lar and Neurosurgical Treatment of Unruptured Intracranial Aneurysms: A Systematic Review and Meta-analysis, JAMA Neurology 76 (3) (2019) 282–293...

work page doi:10.1001/jamaneurol.2018.4165 2019
[7]

B. G. Thompson, R. D. Brown, S. Amin-Hanjani, J. P. Broderick, K. M. Cockroft, E. S. Connolly, G. R. Duckwiler, C. C. Harris, V. J. Howard, S. C. C. Johnston, P. M. Meyers, A. Molyneux, C. S. Ogilvy, A. J. Ringer, J. Torner, on behalf of the American Heart Association Stroke Council, Council on Cardiovascular and Stroke Nursing, and Council on Epidemiolog...

work page doi:10.1161/str.0000000000000070 2015
[8]

Etminan, D

N. Etminan, D. A. d. Sousa, C. Tiseo, R. Bourcier, H. Desal, A. Lind- gren, T. Koivisto, D. Netuka, S. Peschillo, S. L´ emeret, A. Lal, M. D. Vergouwen, G. J. Rinkel, European Stroke Organisation (ESO) guide- lines on management of unruptured intracranial aneurysms, European Stroke Journal 7 (3) (2022) V. doi:10.1177/23969873221099736. 24

work page doi:10.1177/23969873221099736 2022
[9]

Backes, M

D. Backes, M. D. Vergouwen, A. T. Tiel Groenestege, A. S. E. Bor, B. K. Velthuis, J. P. Greving, A. Algra, M. J. Wermer, M. A. van Walderveen, K. G. terBrugge, R. Agid, G. J. Rinkel, PHASES Score for Predic- tion of Intracranial Aneurysm Growth, Stroke 46 (5) (2015) 1221–1226. doi:10.1161/STROKEAHA.114.008198

work page doi:10.1161/strokeaha.114.008198 2015
[10]

Backes, G

D. Backes, G. J. Rinkel, J. P. Greving, B. K. Velthuis, Y. Murayama, H. Takao, T. Ishibashi, M. Igase, K. G. terBrugge, R. Agid, J. E. J¨ a¨ askel¨ ainen, A. E. Lindgren, T. Koivisto, M. Von Und Zu Fraun- berg, S. Matsubara, J. Moroi, G. K. Wong, J. M. Abrigo, K. Igase, K. Matsumoto, M. J. Wermer, M. A. Van Walderveen, A. Algra, M. D. Vergouwen, ELAPSS sc...

work page doi:10.1212/wnl.0000000000003865 2017
[11]

H. J. Kim, D. Y. Yoon, E. S. Kim, H. J. Lee, H. J. Jeon, J. Y. Lee, B.- M. Cho, Intraobserver and interobserver variability in CT angiography and MR angiography measurements of the size of cerebral aneurysms, Neuroradiology 59 (5) (2017) 491–497. doi:10.1007/s00234-017-1826-y

work page doi:10.1007/s00234-017-1826-y 2017
[12]

Timmins, H

K. Timmins, H. Kuijf, M. Vergouwen, M. Otten, Y. Ruigrok, B. Velthuis, I. van der Schaaf, Reliability and Agreement of 2D and 3D Measure- ments on MRAs for Growth Assessment of Unruptured Intracranial Aneurysms, AJNR: American Journal of Neuroradiology 42 (9) (2021) 1598–1603. doi:10.3174/ajnr.A7186

work page doi:10.3174/ajnr.a7186 2021
[13]

Planinc, N

A. Planinc, N. ˇSpegel, Z. Podobnik, U. ˇSinigoj, P. Skubic, J. H. Choi, W. Park, T. Robiˇ c, N. Tabor, L. Jarabek, ˇZ. ˇSpiclin, ˇZ. Bizjak, As- sessing accuracy and consistency in intracranial aneurysm sizing: hu- man expertise vs. artificial intelligence, Scientific Reports 14 (1) (2024) 16080. doi:10.1038/s41598-024-65825-4

work page doi:10.1038/s41598-024-65825-4 2024
[14]

Malhotra, X

A. Malhotra, X. Wu, H. P. Forman, H. K. Grossetta Nardini, C. C. Matouk, D. Gandhi, C. Moore, P. Sanelli, Growth and Rupture Risk of Small Unruptured Intracranial Aneurysms: A Systematic Review, Annals of Internal Medicine 167 (1) (2017) 26–33. doi:10.7326/M17- 0246

work page doi:10.7326/m17- 2017
[15]

Boussel, V

L. Boussel, V. Rayz, C. McCulloch, A. Martin, G. Acevedo- Bolton, M. Lawton, R. Higashida, W. S. Smith, W. L. Young, 25 D. Saloner, Aneurysm Growth Occurs at Region of Low Wall Shear Stress: Patient-Specific Correlation of Hemodynamics and Growth in a Longitudinal Study, Stroke 39 (11) (2008) 2997–3002. doi:10.1161/STROKEAHA.108.521617

work page doi:10.1161/strokeaha.108.521617 2008
[16]

Firouzian, R

A. Firouzian, R. Manniesing, C. T. Metz, S. Klein, B. K. Velthuis, G. J. E. Rinkel, A. Van Der Lugt, W. J. Niessen, Intracranial aneurysm growth quantification in CTA, San Diego, California, USA, 2012, p. 831448. doi:10.1117/12.910713

work page doi:10.1117/12.910713 2012
[17]

X. Liu, H. Haraldsson, Y. Wang, E. Kao, M. Ballweber, A. Mar- tin, C. McCulloch, F. Faraji, D. Saloner, for the UCSF Intracranial Aneurysm Monitoring Group, A Volumetric Metric for Monitoring In- tracranial Aneurysms: Repeatability and Growth Criteria in a Longitu- dinal MR Imaging Study, American Journal of Neuroradiology 42 (9) (2021) 1591–1597. doi:10....

work page doi:10.3174/ajnr.a7190 2021
[18]

Bizjak, ˇZ

ˇZ. Bizjak, ˇZ. ˇSpiclin, Aneurysm growth evaluation and detection: a computer-assisted follow-up MRA analysis, Scientific Reports 14 (1) (2024) 19609. doi:10.1038/s41598-024-70453-z

work page doi:10.1038/s41598-024-70453-z 2024
[19]

Goudarzian, K

F. Goudarzian, K. Kondratiuk, V. L. Rayz, Predicting Cerebral Aneurysm Rupture, Neuroimaging Clinics of North America 35 (3) (2025) 333–347. doi:10.1016/j.nic.2025.05.002

work page doi:10.1016/j.nic.2025.05.002 2025
[20]

J. H. Maki, M. R. Prince, F. J. Londy, T. L. Chenevert, The ef- fects of time varying intravascular signal intensity and k-space ac- quisition order on three-dimensional MR angiography image qual- ity, Journal of Magnetic Resonance Imaging 6 (4) (1996) 642–651. doi:10.1002/jmri.1880060413

work page doi:10.1002/jmri.1880060413 1996
[21]

Tsuruda, D

J. Tsuruda, D. Saloner, D. Norman, Artifacts associated with MR neu- roangiography., American Journal of Neuroradiology 13 (5) (1992) 1411

1992
[22]

W. E. Lorensen, H. E. Cline, Marching cubes: A high resolution 3D sur- face construction algorithm, SIGGRAPH Comput. Graph. 21 (4) (1987) 163–169. doi:10.1145/37402.37422

work page doi:10.1145/37402.37422 1987
[23]

C. M. De Nys, E. S. Liang, M. Prior, M. A. Woodruff, J. I. Novak, A. R. Murphy, Z. Li, C. D. Winter, M. C. Allenby, Time-of-Flight MRA of In- tracranial Aneurysms with Interval Surveillance, Clinical Segmentation 26 and Annotations, Scientific Data 11 (1) (2024) 555. doi:10.1038/s41597- 024-03397-8

work page doi:10.1038/s41597- 2024
[24]

K. A. M. Hackenberg, A. Algra, R. Al-Shahi Salman, J. Fr¨ osen, D. Hasan, S. Juvela, D. Langer, P. Meyers, A. Morita, G. Rinkel, N. Etminan, the Unruptured Aneurysms and SAH CDE Project In- vestigators, J. I. Suarez, R. L. Macdonald, S. Amin-Hanjani, R. D. Brown, A. L. De Oliveira Manoel, C. P. Derdeyn, N. Etminan, E. Keller, P. D. LeRoux, S. Mayer, A. Mo...

work page doi:10.1007/s12028-019-00729-0 2019
[25]

M. D. Hoffman, A. Gelman, The No-U-Turn sampler: Adaptively setting path lengths in hamiltonian monte carlo, Journal of Machine Learning Research 15 (1) (2014) 1593–1623

2014
[26]

A. Hans, S. Bhattacharya, I. Bilionis, P. P. Vlachos, Stochastic volumet- ric reconstruction, in: 15th Int. Symp. on Particle Image Velocimetry- ISPIV, 2023

2023
[27]

A. Hans, S. Bhattacharya, K. Hao, P. Vlachos, I. Bilionis, Bayesian re- construction of 3d particle positions in high-seeding density flows, Mea- surement Science and Technology 35 (11) (2024) 116002. 27

2024
[28]

T. K. Koo, M. Y. Li, A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research, Journal of Chiropractic Medicine 15 (2) (2016) 155–163. doi:10.1016/j.jcm.2016.02.012

work page doi:10.1016/j.jcm.2016.02.012 2016
[29]

A. Hans, A. M. Chaudhari, I. Bilionis, J. H. Panchal, Quantifying in- dividuals’ theory-based knowledge using probabilistic causal graphs: a bayesian hierarchical approach, in: International Design Engineering Technical Conferences and Computers and Information in Engineer- ing Conference, Vol. 83921, American Society of Mechanical Engineers, 2020, p. V003T03A014

2020
[30]

A. Hans, A. M. Chaudhari, I. Bilionis, J. H. Panchal, A bayesian hierar- chical model for extracting individuals’ theory-based causal knowledge, Journal of Computing and Information Science in Engineering 23 (3) (2023) 031011

2023
[31]

Bullitt, D

E. Bullitt, D. Zeng, B. Mortamet, A. Ghosh, S. R. Aylward, W. Lin, B. L. Marks, K. Smith, The effects of healthy ag- ing on intracerebral blood vessels visualized by magnetic reso- nance angiography, Neurobiology of Aging 31 (2) (2010) 290–300. doi:10.1016/j.neurobiolaging.2008.03.022

work page doi:10.1016/j.neurobiolaging.2008.03.022 2010
[32]

M. C. Brindise, S. Rothenberger, B. Dickerhoff, S. Schnell, M. Markl, D. Saloner, V. L. Rayz, P. P. Vlachos, Multi-modality cerebral aneurysm haemodynamic analysis:in vivo4D flow MRI,in vitrovolumet- ric particle velocimetry andin silicocomputational fluid dynamics, Journal of The Royal Society Interface 16 (158) (2019) 20190465. doi:10.1098/rsif.2019.0465

work page doi:10.1098/rsif.2019.0465 2019
[33]

A. Hans, A. Singh, P. Vlachos, I. Bilionis, Smurf: Scalable method for unsupervised reconstruction of flow in 4d flow mri, arXiv preprint arXiv:2505.12494 (2025)

work page arXiv 2025
[34]

W.-C. Hsu, M. Meuschke, A. F. Frangi, B. Preim, K. Lawonn, A survey of intracranial aneurysm detection and segmentation, Medical Image Analysis 101 (2025) 103493. doi:10.1016/j.media.2025.103493. 28

work page doi:10.1016/j.media.2025.103493 2025