pith. sign in

arxiv: 1906.09763 · v1 · pith:DWJA3EEQnew · submitted 2019-06-24 · 📡 eess.IV · cs.CV· physics.med-ph

Improving CCTA based lesions' hemodynamic significance assessment by accounting for partial volume modeling in automatic coronary lumen segmentation

Moti Freiman , Hannes Nickisch , Sven Prevrhal , Holger Schmitt , Mani Vembar , P\'al Maurovich-Horvat , Patrick Donnelly , Liran Goshen This is my paper

Pith reviewed 2026-05-25 17:15 UTC · model grok-4.3

classification 📡 eess.IV cs.CVphysics.med-ph
keywords partial volume effectscoronary lumen segmentationCCTAhemodynamic significanceflow simulationFFRROC analysis
0
0 comments X

The pith

Accounting for partial volume effects in automatic coronary lumen segmentation from CCTA raises flow simulation specificity from 0.6 to 0.68 at fixed sensitivity of 0.83.

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

The paper evaluates whether adding partial volume effect modeling to an automatic algorithm for segmenting coronary lumens from CCTA scans improves the accuracy of downstream computational flow simulations used to judge lesion hemodynamic significance. Segmentation overlap metrics stay comparable to prior methods, yet the flow-based detection of significant coronary disease shows higher specificity and a statistically significant rise in ROC area under the curve from 0.76 to 0.8 when validated against invasive FFR on 76 cases. The authors conclude that partial volume modeling has the potential to reduce false positives in CCTA-based assessments that would otherwise trigger unnecessary invasive exams.

Core claim

Integrating partial volume effect analysis into the automatic coronary lumen segmentation algorithm improves flow simulation specificity from 0.6 to 0.68 while keeping sensitivity at 0.83, and increases the area under the ROC curve for detecting hemodynamically significant CAD from 0.76 to 0.8 against invasive FFR threshold of 0.8.

What carries the argument

Partial volume effect (PVE) modeling added to the automatic coronary lumen segmentation algorithm, which refines the estimated lumen geometry to compensate for CT blurring before the geometry is passed to the flow simulation.

If this is right

  • Fewer lesions diagnosed as obstructive on CCTA would be incorrectly flagged as needing invasive evaluation or revascularization.
  • The overall accuracy of non-invasive CCTA-based hemodynamic assessment increases without changing the underlying flow solver.
  • Segmentation accuracy measured by Dice or surface distance alone does not predict the clinical benefit; downstream flow simulation performance must be checked directly.

Where Pith is reading between the lines

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

  • The same PVE correction step could be inserted into other vessel segmentation pipelines that feed into flow or pressure calculations.
  • The observed gain may vary with scanner resolution or reconstruction kernel; testing on lower-resolution or motion-affected scans would quantify dependence.
  • Independent validation on a multi-center cohort with different CT vendors would establish whether the specificity improvement generalizes beyond the study population.

Load-bearing premise

The computational flow simulation run on the segmented lumen geometry produces values that correctly match real-world hemodynamic significance as measured by invasive FFR.

What would settle it

On the same 76 cases, simulated FFR values derived from the PVE-adjusted segmentations diverge from the invasive FFR measurements in a direction that eliminates or reverses the reported gains in specificity and AUC.

Figures

Figures reproduced from arXiv: 1906.09763 by Hannes Nickisch, Holger Schmitt, Liran Goshen, Mani Vembar, Moti Freiman, P\'al Maurovich-Horvat, Patrick Donnelly, Sven Prevrhal.

Figure 1
Figure 1. Figure 1: The effect of PVE on estimating vessel radius using the full-width half maximum rule19 on 2D vessel profiles with varying stenosis percentage due to the presence of non-calcified plaque. (a)The reference image, including ideal vessel profiles with varying stenosis percentage due to the presence of soft-plaque. (b) The reconstructed image, with the reference segmentation in green and full-width half maximum… view at source ↗
Figure 2
Figure 2. Figure 2: The coronary lumen segmentation algorithm schematic flowchart. The algorithm required the following inputs: 1) the CCTA volume, 2) the coronary artery centerlines, and 3) the aortic root segmentation. The algorithm consists of the following steps: 1) Analysis of the intensity profile along the coronary centerline to detect regions with small diameter lumen that may be overestimated due to the PVE, 2) Estim… view at source ↗
Figure 3
Figure 3. Figure 3: Experimental measurement of the percentage HU reduction as a function of the coronary diameter along with the fitted model (Eq. 4) where HU is the measured HU at the vessel centerline and HU0 is the expected HU at the location without the PVE [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Segmentation performance measures as a function of the two key parameters of [PITH_FULL_IMAGE:figures/full_fig_p021_4.png] view at source ↗
read the original abstract

Purpose: The goal of this study was to assess the potential added benefit of accounting for partial volume effects (PVE) in an automatic coronary lumen segmentation algorithm from coronary computed tomography angiography (CCTA). Materials and methods: We assessed the potential added value of PVE integration as a part of the automatic coronary lumen segmentation algorithm by means of segmentation accuracy using the MICCAI 2012 challenge framework and by means of flow simulation overall accuracy, sensitivity, specificity, negative and positive predictive values and the receiver operated characteristic (ROC) area under the curve. We also evaluated the potential benefit of accounting for PVE in automatic segmentation for flow-simulation for lesions that were diagnosed as obstructive based on CCTA, which could have indicated a need for an invasive exam and revascularization. Results: Our segmentation algorithm improves the maximal surface distance error by ~39% compared to previously published method on the 18 datasets 50 from the MICCAI 2012 challenge with comparable Dice and mean surface distance. Results with and without accounting for PVE were comparable. In contrast, integrating PVE analysis into an automatic coronary lumen segmentation algorithm improved the flow simulation specificity from 0.6 to 0.68 with the same sensitivity of 0.83. Also, accounting for PVE improved the area under the ROC curve for detecting hemodynamically significant CAD from 0.76 to 0.8 compared to automatic segmentation without PVE analysis with invasive FFR threshold of 0.8 as the reference standard. The improvement in the AUC was statistically significant (N=76, Delong's test, p=0.012). Conclusion: Accounting for the partial volume effects in automatic coronary lumen segmentation algorithms has the potential to improve the accuracy of CCTA-based hemodynamic assessment of coronary artery lesions.

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.

Referee Report

2 major / 2 minor

Summary. The manuscript describes an automatic coronary lumen segmentation algorithm from CCTA that incorporates partial volume effect (PVE) modeling. Segmentation performance is evaluated on the MICCAI 2012 challenge dataset (18 cases), showing a ~39% reduction in maximal surface distance error with comparable Dice and mean surface distance scores with/without PVE. Downstream, flow simulations on N=76 cases demonstrate improved specificity (0.6 to 0.68 at sensitivity 0.83) and AUC (0.76 to 0.8, Delong p=0.012) for detecting hemodynamically significant CAD (FFR threshold 0.8) when PVE is included.

Significance. If the reported gains hold under full methodological disclosure, the work could meaningfully improve CCTA-based non-invasive assessment of lesion hemodynamic significance by refining lumen geometry inputs to flow models, with potential to reduce false positives and unnecessary invasive exams. The use of an external challenge dataset for segmentation and independent invasive FFR as reference for the ROC analysis provides a reasonably grounded evaluation framework.

major comments (2)
  1. [Materials and methods] Materials and methods: The exact segmentation algorithm equations, PVE correction formulation, and integration steps are not specified. This is load-bearing for the central claim because the abstract states that Dice/mean surface distance remain comparable with/without PVE while flow-simulation specificity and AUC improve; without the equations it is impossible to determine whether the downstream gains arise from genuine lumen-geometry refinement or from unstated interactions between the PVE adjustment and the flow model.
  2. [Results] Results / flow simulation: No description is given of the flow model (boundary conditions, outlet resistances, steady vs. pulsatile assumptions, or rigid-wall approximation). Because the headline result (specificity 0.6→0.68, AUC 0.76→0.8) is obtained entirely through this model applied to the segmented geometries, the absence of these details leaves open the possibility that the statistically significant metric improvements reflect model-specific sensitivity to the small lumen-area changes induced by PVE rather than improved clinical assessment of hemodynamic significance.
minor comments (2)
  1. [Abstract] Abstract: The phrase “18 datasets 50 from the MICCAI 2012 challenge” appears to contain a typographical error and should be clarified.
  2. [Results] The manuscript does not report error bars, confidence intervals, or cross-validation details for the segmentation or flow-simulation metrics, which would aid interpretation of the reported numeric gains.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive feedback. The comments highlight important gaps in methodological transparency that we will address in revision. Below we respond point-by-point to the major comments.

read point-by-point responses

  1. Referee: [Materials and methods] The exact segmentation algorithm equations, PVE correction formulation, and integration steps are not specified. This is load-bearing for the central claim because the abstract states that Dice/mean surface distance remain comparable with/without PVE while flow-simulation specificity and AUC improve; without the equations it is impossible to determine whether the downstream gains arise from genuine lumen-geometry refinement or from unstated interactions between the PVE adjustment and the flow model.

    Authors: We agree that the absence of the explicit equations and integration steps limits the ability to interpret the source of the observed improvements. In the revised manuscript we will add the full mathematical formulation of the automatic segmentation algorithm, the partial-volume-effect correction term, and the precise manner in which the corrected lumen surface is passed to the flow solver. This addition will make it possible to verify that the specificity and AUC gains originate from refined lumen geometry rather than from any hidden interaction with the downstream model. revision: yes

  2. Referee: [Results] Results / flow simulation: No description is given of the flow model (boundary conditions, outlet resistances, steady vs. pulsatile assumptions, or rigid-wall approximation). Because the headline result (specificity 0.6→0.68, AUC 0.76→0.8) is obtained entirely through this model applied to the segmented geometries, the absence of these details leaves open the possibility that the statistically significant metric improvements reflect model-specific sensitivity to the small lumen-area changes induced by PVE rather than improved clinical assessment of hemodynamic significance.

    Authors: We acknowledge that a complete description of the flow-simulation pipeline is required to substantiate the clinical relevance of the reported gains. The revised manuscript will include the boundary conditions, outlet resistance assignment method, steady-flow assumption, and rigid-wall approximation employed. With these details provided, readers will be able to assess whether the specificity and AUC improvements are attributable to more accurate lumen geometry or to particular sensitivities of the chosen flow model. revision: yes

Circularity Check

0 steps flagged

No significant circularity; central claims rest on external validation against MICCAI dataset and independent invasive FFR

full rationale

The paper's derivation chain consists of (1) automatic lumen segmentation with/without PVE modeling, (2) application of a flow simulation model to the resulting geometries, and (3) direct comparison of simulated hemodynamics to invasive FFR ground truth on an external cohort of 76 lesions. Segmentation accuracy is benchmarked on the public MICCAI 2012 challenge set; flow-simulation metrics (specificity 0.6→0.68, AUC 0.76→0.8, Delong p=0.012) are likewise computed against independent FFR measurements. No equation, fitted parameter, or self-citation reduces these reported gains to quantities defined inside the paper itself. The flow model is treated as an external black-box predictor whose validity is an assumption, not a tautology internal to the derivation.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard assumptions of medical image processing and computational fluid dynamics rather than new postulates; no invented physical entities are introduced.

free parameters (1)
  • PVE correction parameters
    The automatic segmentation algorithm incorporates tunable parameters for modeling partial volume effects whose exact values are not stated in the abstract.
axioms (2)
  • domain assumption Partial volume effects in CCTA can be modeled to improve lumen boundary estimation
    This modeling choice is the explicit intervention whose benefit is being measured.
  • domain assumption Flow simulation on segmented lumen geometry predicts hemodynamic significance
    The paper treats the simulation output as a proxy for invasive FFR without deriving or validating the fluid model inside the manuscript.

pith-pipeline@v0.9.0 · 5901 in / 1454 out tokens · 36777 ms · 2026-05-25T17:15:26.691746+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages

  1. [1]

    the ratio between the pressure after a lesion and the normal pressure) can be assessed from CCTA data using flow simulations

    Introduction: Coronary artery disease (CAD) is the single leading cause of death worldwide, accounting for 11.2% of all deaths globally in 2011.1 Among the non-invasive tests available for patients with suspected CAD, Coronary Computed Tomography Angiography (CCTA) is a rapidly evolving technique to rule out CAD due to its high negative predictive value.2...

    work page 2011

  2. [2]

    Datasets: We used two data sets as follows

    Materials and methods: 2.A. Datasets: We used two data sets as follows. The first dataset was the publicly available MICCAI 2012 coronary artery segmentation challenge database. The database consists of 48 CCTA datasets that were acquired from a representative selection of CAD symptomatic patients using several cardiac CT scanners from different vendors w...

    work page 2012

  3. [3]

    The coronary-artery centerlines

    work page

  4. [4]

    The segmentation of the aortic root 10 The coronary artery centerlines and the aorta segmentation were computed automatically and adjusted manually by a cardiac CT expert (M.V) to account for algorithm inaccuracies using a commercially available software dedicated for cardiac image analysis (Comprehensive Cardiac Analysis, IntelliSpace Portal 6.0, Philips...

    work page

  5. [5]

    Evaluation 3.A. Methodology: We implemented our main algorithm in C++ using the graph min-cut solver of Boykov et al25, and an accelerated approximate K-nearest neighbor search.24 We experimentally set the value of the regularization term 𝜆 in Eq. 5 to 1.75 and 𝐾 in Eq. 6 to 100. The average running time to segment the entire coronary tree lumen for each ...

    work page 2012

  6. [6]

    and 𝐾 (Eq. 6). We assessed the sensitivity of our algorithm to the two key parameters in our algorithm by using the training data available from the MICCAI 2012 challenge dataset. 12 3.A.3. Impact of accounting for PVE on simulated FFR performance Next, we assessed the performance of simulated FFR measurements based on automatically generated coronary 3D ...

    work page 2012

  7. [7]

    Quantitative analysis of the CAD from CCTA required both automatic extraction of the coronaries’ centerlines30–34 and automatic segmentation of the coronaries lumen

    Discussion: Our study demonstrates the importance of accounting for PVE in automatic coronary segmentation algorithms used to determine the hemodynamic significance of coronary artery stenosis by CCTA based on flow simulations. Quantitative analysis of the CAD from CCTA required both automatic extraction of the coronaries’ centerlines30–34 and automatic s...

    work page 2012

  8. [8]

    2014 Global geographic analysis of mortality from ischaemic heart disease by country, age and income: Statistics from World Health Organisation and United Nations

    Nowbar AN, Howard JP, Finegold JA, Asaria P, Francis DP. 2014 Global geographic analysis of mortality from ischaemic heart disease by country, age and income: Statistics from World Health Organisation and United Nations. Int J Cardiol. 2014;174(2):293-298. doi:10.1016/j.ijcard.2014.04.096

    work page doi:10.1016/j.ijcard.2014.04.096 2014

  9. [9]

    CT coronary angiography in patients with suspected angina due to coronary heart disease (SCOT-HEART): an open-label, parallel-group, multicentre trial

    Scot-heart T. CT coronary angiography in patients with suspected angina due to coronary heart disease (SCOT-HEART): an open-label, parallel-group, multicentre trial. Lancet. 2015;385(9985):2383-2391. doi:10.1016/S0140- 6736(15)60291-4

    work page doi:10.1016/s0140- 2015

  10. [10]

    Comprehensive Assessment of Coronary Artery Stenoses

    Meijboom WB, Van Mieghem CAG, van Pelt N, et al. Comprehensive Assessment of Coronary Artery Stenoses. Computed Tomography Coronary Angiography Versus Conventional Coronary Angiography and Correlation With Fractional Flow Reserve in Patients With Stable Angina. J Am Coll Cardiol. 2008;52(8):636-643. doi:10.1016/j.jacc.2008.05.024

    work page doi:10.1016/j.jacc.2008.05.024 2008

  11. [11]

    Gonzalez JA, Lipinski MJ, Flors L, Shaw PW, Kramer CM, Salerno M. Meta- analysis of diagnostic performance of coronary computed tomography angiography, computed tomography perfusion, and computed tomography- fractional flow reserve in functional myocardial ischemia assessment versus invasive fractional flow reserve. Am J Cardiol. 2015;116(9):1469-1478. do...

    work page doi:10.1016/j.amjcard.2015.07.078 2015

  12. [12]

    Diagnostic performance of non-invasive fractional flow reserve derived from coronary CT angiography in suspected coronary artery disease: The NXT trial

    Nørgaard BL, Leipsic J, Gaur S, et al. Diagnostic performance of non-invasive fractional flow reserve derived from coronary CT angiography in suspected coronary artery disease: The NXT trial. J Am Coll Cardiol. 2014;63(12):1145- 33

    work page 2014

  13. [13]

    doi:10.1016/j.jacc.2013.11.043

    work page doi:10.1016/j.jacc.2013.11.043 2013

  14. [14]

    Noninvasive fractional flow reserve derived from coronary CT angiography clinical data and scientific principles

    Min JK, Taylor CA, Achenbach S, et al. Noninvasive fractional flow reserve derived from coronary CT angiography clinical data and scientific principles. JACC Cardiovasc Imaging. 2015;8(10):1209-1222. doi:10.1016/j.jcmg.2015.08.006

    work page doi:10.1016/j.jcmg.2015.08.006 2015

  15. [15]

    Diagnostic accuracy of fractional flow reserve from anatomic CT angiography

    Min JK, Leipsic J, Pencina MJ, et al. Diagnostic accuracy of fractional flow reserve from anatomic CT angiography. Jama. 2012;308(12):1237-1245. doi:10.1001/2012.jama.11274

    work page doi:10.1001/2012.jama.11274 2012

  16. [16]

    Fractional flow reserve computed from noninvasive CT angiography data: diagnostic performance of an on-site clinician- operated computational fluid dynamics algorithm

    Coenen A, Lubbers MM, Kurata A, et al. Fractional flow reserve computed from noninvasive CT angiography data: diagnostic performance of an on-site clinician- operated computational fluid dynamics algorithm. Radiology. 2015;274(3):674-

    work page 2015

  17. [17]

    doi:10.1148/radiol.14140992

    work page doi:10.1148/radiol.14140992

  18. [18]

    Kirbas, C. Quek F. A review of vessel extraction techniques and algorithms. Comput Surv. 2004;36(2):81-121. doi:10.1145/1031120.1031121

    work page doi:10.1145/1031120.1031121 2004

  19. [19]

    A review of 3D vessel lumen segmentation techniques: Models, features and extraction schemes

    Lesage D, Angelini ED, Bloch I, Funka-Lea G. A review of 3D vessel lumen segmentation techniques: Models, features and extraction schemes. Med Image Anal. 2009;13(6):819-845. doi:10.1016/j.media.2009.07.011

    work page doi:10.1016/j.media.2009.07.011 2009

  20. [20]

    Standardized evaluation framework for evaluating coronary artery stenosis detection, stenosis quantification and lumen segmentation algorithms in computed tomography angiography

    Kirişli HA, Schaap M, Metz CT, et al. Standardized evaluation framework for evaluating coronary artery stenosis detection, stenosis quantification and lumen segmentation algorithms in computed tomography angiography. Med Image Anal. 2013;17(8):859-876. doi:10.1016/j.media.2013.05.007

    work page doi:10.1016/j.media.2013.05.007 2013

  21. [21]

    http://coronary.bigr.nl/stenoses/

    Coronary Artery Stenoses Detection and Quantification Evaluation Framework. http://coronary.bigr.nl/stenoses/. 34

    work page

  22. [22]

    Globally optimal segmentation of interacting surfaces with geometric constraints

    Li KLK, Wu XWX, Chen DZ, Sonka M. Globally optimal segmentation of interacting surfaces with geometric constraints. Proc 2004 IEEE Comput Soc Conf Comput Vis Pattern Recognition, 2004 CVPR 2004. 2004;1:0-5. doi:10.1109/CVPR.2004.1315059

    work page doi:10.1109/cvpr.2004.1315059 2004

  23. [23]

    Bayesian Segmentation of Atrium Wall Using Globally-Optimal Graph Cuts on 3D Meshes

    Veni G, Fu Z, Awate SP, Whitaker RT. Bayesian Segmentation of Atrium Wall Using Globally-Optimal Graph Cuts on 3D Meshes. In: Gee JC, Sarang J, Kilian MP, Wells WM, Zöllei L, EDS. Inf Process Med Imaging. LNCS, Vol. 7917. Springer-Verlag Berlin, Heidelberg; 2013:656–667. doi: 10.1007/978-3-642- 38868-2_55

    work page doi:10.1007/978-3-642- 2013

  24. [24]

    Improving accuracy in coronary lumen segmentation via explicit calcium exclusion, learning-based ray detection and surface optimization

    Lugauer F, Zhang J, Zheng Y, Hornegger J, Kelm BM. Improving accuracy in coronary lumen segmentation via explicit calcium exclusion, learning-based ray detection and surface optimization. SPIE Med Imaging. 2014;9034:90343U- 90343U-10. doi:10.1117/12.2043238

    work page doi:10.1117/12.2043238 2014

  25. [25]

    Precise lumen segmentation in coronary computed tomography angiography

    Lugauer F, Zheng Y, Hornegger J, Kelm BM. Precise lumen segmentation in coronary computed tomography angiography. In: Menze B, Langs G, Montillo A, et al, eds. Medical Computer Vision: Algorithms for Big Data LNCS, Vol. 8848. Springer-Verlag Berlin, Heidelberg; 2014:137–147. doi:10.1007/978-3-319-13972- 2_13

    work page doi:10.1007/978-3-319-13972- 2014

  26. [26]

    Accurate Quantification of Small-Diameter Tubular Structures in Isotropic CT Volume Data Based on Multiscale Line Filter Responses

    Sato Y, Yamamoto S, Tamura S. Accurate Quantification of Small-Diameter Tubular Structures in Isotropic CT Volume Data Based on Multiscale Line Filter Responses. In: Barillot C, Haynor DR, and Hellier P, eds. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2004: 7th International Conference, LNCS, Vol. 3216. Springer International Pub...

    work page doi:10.1007/978-3-540-30135-6_62 2004

  27. [27]

    Accuracy of CT-based thickness measurement of thin structures: modeling of limited spatial resolution in all three dimensions

    Prevrhal S, Fox JC, Shepherd JA, Genant HK. Accuracy of CT-based thickness measurement of thin structures: modeling of limited spatial resolution in all three dimensions. Med Phys. 2003;30(1):1-8. doi:10.1118/1.1521940

    work page doi:10.1118/1.1521940 2003

  28. [28]

    Full width at half maximum as a measure of vessel diameter in computed tomography angiography

    Varma JK, Subramanyan K, Durgan J. Full width at half maximum as a measure of vessel diameter in computed tomography angiography. In: Vol 5372. ; 2004:447-454. http://dx.doi.org/10.1117/12.535642

    work page doi:10.1117/12.535642 2004

  29. [29]

    Automatic coronary lumen segmentation with partial volume modeling improves lesions’ hemodynamic significance assessment

    Freiman M, Lamash Y, Gilboa G, et al. Automatic coronary lumen segmentation with partial volume modeling improves lesions’ hemodynamic significance assessment. In: SPIE Medical Imaging. ; 2016:978403. http://dx.doi.org/10.1117/12.2209476

    work page doi:10.1117/12.2209476 2016

  30. [30]

    Computer- aided simple triage (CAST) for coronary CT angiography (CCTA)

    Goldenberg R, Eilot D, Begelman G, Walach E, Ben-Ishai E, Peled N. Computer- aided simple triage (CAST) for coronary CT angiography (CCTA). Int J Comput Assist Radiol Surg. 2012;7(6):819-827. doi:10.1007/s11548-012-0684-7

    work page doi:10.1007/s11548-012-0684-7 2012

  31. [31]

    Automatic segmentation, detection and quantification of coronary artery stenoses on CTA

    Shahzad R, Kirişli H, Metz C, et al. Automatic segmentation, detection and quantification of coronary artery stenoses on CTA. Int J Cardiovasc Imaging. 2013:1847-1859. doi:10.1007/s10554-013-0271-1

    work page doi:10.1007/s10554-013-0271-1 2013

  32. [32]

    Nearest neighbor pattern classification

    Cover T, Hart P. Nearest neighbor pattern classification. IEEE Trans Inf Theory. 1967;13(1):21-27. doi:10.1109/TIT.1967.1053964

    work page doi:10.1109/tit.1967.1053964 1967

  33. [33]

    Fast Approximate Nearest Neighbors with Automatic Algorithm Configuration

    Muja M, Lowe DG. Fast Approximate Nearest Neighbors with Automatic Algorithm Configuration. Int Conf Comput Vis Theory Appl (VISAPP ’09). 2009:1-10. doi:10.1.1.160.1721

    work page 2009

  34. [34]

    Graph cuts and efficient N-D image segmentation

    Boykov Y, Funka-Lea G. Graph cuts and efficient N-D image segmentation. Int J 36 Comput Vis. 2006;70(2):109-131. doi:10.1007/s11263-006-7934-5

    work page doi:10.1007/s11263-006-7934-5 2006

  35. [35]

    Learning Patient-Specific Lumped Models for Interactive Coronary Blood Flow Simulations

    Nickisch H, Lamash Y, Prevrhal S, et al. Learning Patient-Specific Lumped Models for Interactive Coronary Blood Flow Simulations. In: Navab N, Hornegger J, Wells MW, Frangi FA, eds. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015: 18th International Conference, LNCS, Vol. 9350. Springer-Verlag Berlin, Heidelberg; 2015:433-441. do...

    work page doi:10.1007/978-3-319-24571- 2015

  36. [36]

    Intraspecific scaling laws of vascular trees

    Huo Y, Kassab GS. Intraspecific scaling laws of vascular trees. J R Soc Interface. 2012;9(66):190-200. doi:10.1098/rsif.2011.0270

    work page doi:10.1098/rsif.2011.0270 2012

  37. [37]

    Workstation-Based Calculation of CTA- Based FFR for Intermediate Stenosis

    Kruk M, Wardziak Ł, Demkow M, et al. Workstation-Based Calculation of CTA- Based FFR for Intermediate Stenosis. JACC Cardiovasc Imaging. 2016;9(6):690-

    work page 2016

  38. [38]

    doi:10.1016/j.jcmg.2015.09.019

    work page doi:10.1016/j.jcmg.2015.09.019 2015

  39. [39]

    Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach

    Delong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837-845. doi:10.2307/2531595

    work page doi:10.2307/2531595 1988

  40. [40]

    Segmentation of the heart and great vessels in CT images using a model-based adaptation framework

    Ecabert O, Peters J, Walker MJ, et al. Segmentation of the heart and great vessels in CT images using a model-based adaptation framework. Med Image Anal. 2011;15(6):863-876. doi:10.1016/j.media.2011.06.004

    work page doi:10.1016/j.media.2011.06.004 2011

  41. [41]

    Automatic model-based segmentation of the heart in CT images

    Ecabert O, Peters J, Schramm H, et al. Automatic model-based segmentation of the heart in CT images. IEEE Trans Med Imaging. 2008;27(9):1189-1202. doi:10.1109/TMI.2008.918330

    work page doi:10.1109/tmi.2008.918330 2008

  42. [42]

    Carotid vasculature modeling from patient CT angiography studies for interventional procedures simulation

    Freiman M, Joskowicz L, Broide N, et al. Carotid vasculature modeling from patient CT angiography studies for interventional procedures simulation. Int J 37 Comput Assist Radiol Surg. 2012;7(5):799-812. doi:10.1007/s11548-012-0673-x

    work page doi:10.1007/s11548-012-0673-x 2012

  43. [43]

    Robust and accurate coronary artery centerline extraction in CTA by combining model-driven and data-driven approaches

    Zheng Y, Tek H, Funka-Lea G. Robust and accurate coronary artery centerline extraction in CTA by combining model-driven and data-driven approaches. In: Mori K, Sakuma I, Sato Y, Barillot C, and Navab N, eds. Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2013: 16th International Conference, LNCS, Vol. 8151. Springer-Verlag Berlin, He...

    work page doi:10.1007/978-3-642-40760-4_10 2013

  44. [44]

    Standardized evaluation methodology and reference database for evaluating coronary artery centerline extraction algorithms

    Schaap M, Metz CT, van Walsum T, et al. Standardized evaluation methodology and reference database for evaluating coronary artery centerline extraction algorithms. Med Image Anal. 2009;13(5):701-714. doi:10.1016/j.media.2009.06.003

    work page doi:10.1016/j.media.2009.06.003 2009

  45. [45]

    Computational fluid dynamics applied to cardiac computed tomography for noninvasive quantification of fractional flow reserve: Scientific basis

    Taylor CA, Fonte TA, Min JK. Computational fluid dynamics applied to cardiac computed tomography for noninvasive quantification of fractional flow reserve: Scientific basis. J Am Coll Cardiol. 2013;61(22):2233-2241. doi:10.1016/j.jacc.2012.11.083

    work page doi:10.1016/j.jacc.2012.11.083 2013