Pith. sign in

REVIEW 3 major objections 6 minor 30 references

Self-Calibrated Epipolar Reconstruction for Assessment of Aneurysms in the Internal Carotid Artery Using In-Silico Biplane Angiograms

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper establishes a fully automatic, calibration-object-free epipolar reconstruction pipeline that recovers 3D internal-carotid-artery and aneurysm geometry from simulated biplane angiograms in about 10 seconds, with Dice-Sorensen…

desk verdict Solid, honest in-silico demonstration of a fast self-calibrated epipolar reconstruction pipeline for ICA aneurysms, but the main calibration assumption is untested on real gantry angulation and validation is lean (3 models, no variance). read the letter →

arxiv 2501.11793 v1 pith:U7VPE6NK submitted 2025-01-21 physics.med-ph

Kyle A. Williams , S.V. Setlur Nagesh , Daniel R. Bednarek , Stephen Rudin , Ciprian N. Ionita This is my paper · ORCID
classification physics.med-ph
keywords biplaneangiographyepipolarreconstructionintracranialaneurysminternalcarotidarteryself-calibrationDice-Sorensencoefficientcomputationalfluiddynamics3Dvessel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that a fully automatic, calibration-object-free epipolar reconstruction can turn standard biplane angiograms into a usable 3D view of the internal carotid artery and its aneurysm. It establishes the method on three simulated patient-derived geometries, generating virtual angiograms with CFD and cone-beam projection, and reports Dice-Sorensen coefficients of 0.745, 0.759, and 0.654 against the ground-truth models. The point of the work is that 3D vessel morphology, useful for aneurysm assessment and device sizing, does not require CBCT or an external calibration object; it can be derived from the routine two-view images themselves, with average reconstruction time around 10 seconds. A sympathetic reader would take the paper as showing that the main obstacle to intraprocedural 3D guidance is not geometry recovery but bolus and timing assumptions plus vessel-overlap disambiguation.

What carries the argument

The load-bearing mechanism is the contrast-bolus self-calibration step: because biplane systems share one axial axis, the axial advance of contrast between two time frames should be identical in both views after pixel pitch equalization, so the ratio of measured advances yields the magnification scaling factor k, and the residual distance from the image bottom to the bolus top gives field-of-view alignment. This removes the calibration object and makes all downstream steps operate on geometrically consistent images: essential-matrix estimation via the OpenCV solver, back-projection intersection into bounding boxes, structural projection filtering, and ellipse fitting driven by single-integral line-profile areas.

What would settle it

Take a biplane sequence of a phantom with known vessel geometry and tilt one gantry cranially; if the reconstructed volume's Dice coefficient against the known geometry falls markedly compared with the non-tilted case, the shared-axis bolus assumption is the limiting step, not the triangulation. A direct measurement of the ratio of bolus advance distances in the two views across the sequence would show whether the ratio stays near one as assumed.

Watch

Extended reading notes

Core claim

The central claim is that the two views of a biplane angiogram contain enough self-referential information to calibrate each other: by matching the axial distance a contrast bolus travels between corresponding time points, the algorithm equalizes pixel size and field of view across imagers without knowing magnification a priori. With the views scaled and aligned, feature matching supplies the essential matrix, back-projection of vessel masks yields per-slice bounding boxes, structural projection filtering removes false candidates using temporal maximum-intensity-projection thickness estimates, and a parametric ellipse fit, constrained by line-profile area integration, refines each cross-section. The result is a non-iterative, roughly 10-second 3D reconstruction that the paper validates on three in-silico aneurysm models with Dice coefficients 0.745, 0.759, and 0.654, with error concentrated where vessels overlap in projection.

Load-bearing premise

The method assumes the contrast dye moves the same distance along the shared head-to-foot direction in both X-ray views over the same time interval, which breaks if either imaging arm is angled toward the head or feet.

Editorial extensions

If this is right

  • Standard biplane angiography, without CBCT or a calibration object, can produce a 3D vessel estimate accurate enough to show the aneurysm dome and parent vessel, with Dice coefficients of 0.745, 0.759, and 0.654 in simulation.
  • Because the reconstruction is non-iterative and rule-based, it runs in about 10 seconds, making intraprocedural use during neurointervention plausible.
  • The main failure mode is partial vessel overlap in projection, so clinical application would target vasculature clearly visible in both views or require an overlap-resolving refinement.
  • The method could reduce the need for CTA acquisitions and patient transfers between imaging suites by deriving 3D morphology from routine biplane images.
  • The 3D reconstruction could serve as input for quantitative angiography and device-sizing analyses that require depth information.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Inference: if the self-calibration transfers to clinical frame rates near 3 fps, iterative magnification estimation would likely be needed because bolus advancement between frames becomes coarser; this is a testable extension the authors note as future work.
  • Inference: the bolus-transit equality could be adapted to single-plane rotational angiography by using cardiac phase-gated frames as pseudo-second views, though the shared-axis assumption would need re-derivation.
  • Inference: the reported Dice values on simulated data set an upper bound for clinical performance; real-image quantum mottle, patient motion, and incomplete contrast filling will likely degrade overlap disambiguation before they degrade the geometry solver.
  • Inference: the ellipse-fitting with line-profile area could be extended to non-elliptical cross-sections or to estimate lumen area for flow quantification, replacing the closed-form ellipse solve with the same bounding-box-and-area principle.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper introduces a fully automatic, calibration-object-free epipolar reconstruction pipeline for three-dimensional visualization of internal carotid artery (ICA) aneurysms from biplane angiograms. The method equalizes magnification and field of view between the two views by assuming that the axial advancement of a contrast bolus over a fixed time interval is identical in both projections, then performs feature matching, essential-matrix estimation, back-projection, and ellipse fitting to produce a 3D vessel mask. The pipeline is evaluated on three CFD-derived virtual angiograms generated from patient-specific aneurysm geometries, with reported Dice-Sorensen coefficients of 0.745, 0.759, and 0.654 and an average reconstruction time of about 10 seconds. The authors conclude that the method enhances 3D visualization from routine biplane data and could reduce the need for additional CTA or CBCT acquisitions.

Significance. If the central claim is correct, the proposed method could provide a practical, intraprocedural 3D reconstruction capability for neurointerventional suites without requiring calibration objects or additional imaging. The in-silico validation against exact CFD ground truth geometries is a legitimate proof-of-concept design, and the described pipeline is explainable and non-iterative, which is appealing for clinical translation. However, the study is preliminary: it tests only three models, reports no variance or error bars, lacks a comparison with existing reconstruction methods, and its central self-calibration assumption is violated by routine gantry angulations. The reported Dice values should therefore be interpreted as a feasibility demonstration under idealized geometry rather than evidence of robustness to clinical conditions.

major comments (3)
  1. [Section 2.2; Discussion, fourth paragraph] The self-calibration step assumes that the axial advancement of contrast is identical in both views, which requires the two imagers to share a common axis without relative cranial/caudal angulation. The authors explicitly acknowledge in the Discussion that 'cranial or caudal angulation in either gantry may introduce bias in the magnification or FOV equalization steps' and suggest affine transformations as future work, but no correction is implemented or tested. Since such angulation is routine in clinical biplane neuroangiography, the reported Dice coefficients from common-axis simulations do not support the abstract's claim that the method works on 'standard, routinely acquired biplane angiographic data.' This is a load-bearing external-validity limitation; the authors must either implement and validate a correction for gantry angulation or substantially narrow the claims to common-axis acquisitions and justify that such acquisitions are representative of the intended clinical setting.
  2. [Section 2.4, final paragraph; Section 3.1] The Dice-Sorensen evaluation protocol is not described. After self-calibration, the reconstruction is expressed in the rescaled image pixel grid, but the method does not determine an absolute physical scale because magnification is only equalized between views, not determined in absolute units. The manuscript does not explain how the reconstructed mask is aligned, scaled, and resampled to the CFD ground-truth volume in order to compute the reported Dice values (0.745, 0.759, 0.654). Without specifying the voxel grid, coordinate registration procedure, and binarization threshold, these numeric results cannot be reproduced or interpreted. This is a load-bearing methodological gap because the Dice coefficients are the primary evidence supporting the central claim.
  3. [Section 3.1] The claim that the method 'generalized well across the three tested aneurysm models' is weakly supported. Only three models are tested, with no variance measures, no repeated runs, no sensitivity analysis, and no morphological description of the aneurysm geometries (e.g., size, neck width, tortuosity, degree of vessel overlap). Additionally, no comparison is made to existing epipolar reconstruction methods or to an alternative 3D technique such as 3D DSA or CBCT. At minimum, the limitations of a three-model proof-of-concept should be explicitly acknowledged, and the generalizability claim should be tempered accordingly.
minor comments (6)
  1. [Section 2.2, Figures 2 and 3] The definition of the scaling factor k is ambiguous: the text states the primed system is rescaled by the ratio of transit distances l and l', but it is not clear whether k = l/l' or k = l'/l. Please state the direction explicitly and ensure the rescaling operation in the text matches the figures.
  2. [Section 2.4, Eq. (7)] The expressions for the semi-major and semi-minor axes a and b appear garbled by typesetting and are difficult to verify. Please rewrite Eq. (7) with unambiguous notation and check the algebra, as this equation is central to the ellipse-fitting refinement.
  3. [Abstract and Section 2.1] The abstract states that the method uses 'standard, routinely acquired biplane angiographic data,' while the simulations are performed at an effective 1000 fps frame rate (1 ms time steps). The Discussion notes that the method may be difficult at clinical frame rates such as 3 fps. Please specify the simulated frame rate in the Methods and qualify the abstract's applicability claim accordingly.
  4. [Section 2.3] The use of the essential matrix solver to estimate rotation and translation is described only briefly. The essential matrix yields the relative pose up to an unknown scale; the manuscript does not explain how this scale is determined for back-projection into a common volume. Please clarify whether absolute scale is used or whether the reconstruction is generated in an unscaled coordinate frame.
  5. [Section 3.1, Figures 8 and 9] The captions for Figures 8 and 9 do not describe which model or axial slices are shown, nor the color coding for true positives, false positives, and false negatives in a consistent manner. Please make the captions self-contained and define all color labels explicitly.
  6. [Discussion, Eq. (12)] The equation for x-ray attenuation is numbered as Eq. (12), but the previous numbered equations in Section 2.4 stop at Eq. (8). Please renumber the equations sequentially throughout the manuscript.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; self-calibration uses independent bolus-transit observables and ground-truth geometry is used only for evaluation.

full rationale

The derivation chain is self-contained and not circular. The self-calibration factor k is estimated from the measured contrast bolus transit distances l and l' in the two views (Section 2.2), which are direct image observables and do not presuppose the 3D reconstruction. The epipolar geometry is then computed from matched axial features using a standard essential-matrix solver, and the projection data are used for consistency-based refinement via structural projection filtering and ellipse fitting. The cross-sectional area used in ellipse fitting is obtained by integrating projection line profiles (Eq. 2), which is a data-driven estimate rather than a quantity derived from the output. The CFD ground-truth models are used exclusively to compute Dice-Sorensen coefficients for evaluation; no ground-truth geometry enters the reconstruction pipeline. The paper's own Discussion identifies the common-axis assumption for contrast transit as a potential source of bias under cranial or caudal gantry angulation, but this is an external-validity limitation, not a circular reduction of the reconstruction to its inputs. Existing self-citations (e.g., Ref. 12 for CFD simulation parameters and Refs. 18-23 for high-speed angiography context) are minor and not load-bearing for the central reconstruction claim. No fitted parameter is renamed as a prediction, and no author-imported uniqueness theorem is invoked to force the method choice.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the imaging geometry and shape assumptions listed above; these are standard in angiography reconstruction but each one is an idealized condition that real biplane data will violate to some degree. The method has no free parameters fitted to the reported Dice scores, but it does have unspecified implementation thresholds that act as hand-chosen values.

free parameters (2)
  • Structure elimination thresholds = not specified
    In Section 2.3, the structural projection filter eliminates candidate bounding boxes when the estimated number of structures is 'too many or too few' or the structure is 'much larger' than projected; these thresholds are chosen by hand and not reported, and they affect the Dice score.
  • Ellipse orientation continuity weight = not specified
    Section 2.4 chooses the ellipse orientation that 'most closely matches' the previous slice, but the matching metric and any weighting are not defined; this choice affects the final reconstruction shape.
assumptions (4)
  • domain assumption The two biplane views share a common axis in the axial direction, so the contrast bolus advance over a fixed time interval is equal in both views after pixel-pitch scaling.
    Section 2.2, Figure 2. The self-calibration is built on this equality. The authors note in the Discussion that cranial or caudal gantry angulation would bias this correction, so the assumption is load-bearing and not always true in the clinic.
  • domain assumption Vessel cross-sections can be well approximated by ellipses.
    Section 2.4, citing references 16 and 17. The ellipse fitting refinement assumes this shape model; bifurcations and overlapping vessels violate it in places, as the authors acknowledge.
  • domain assumption The integral of the projection line profile (Eq. 2) approximates the true cross-sectional area of the vessel.
    Eq. 2 in Section 2.4. This holds exactly only for pure depth integration of a binary contrast distribution and ignores exponential x-ray attenuation; the authors acknowledge in the Discussion that a log-corrected algorithm is needed for real data.
  • domain assumption The CFD-simulated angiograms faithfully represent clinical biplane angiograms.
    Section 2.1. The entire validation rests on simulated data that lacks quantum mottle, scatter, patient motion, and polychromatic attenuation, limitations the Discussion lists.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Self-Calibrated Epipolar Reconstruction for Assessment of Aneurysms in the Internal Carotid Artery Using In-Silico Biplane Angiograms." pith.science (2026). https://pith.science/paper/U7VPE6NK

@misc{pith2026250111793,
  author       = {Pith},
  title        = {Pith review of: Self-Calibrated Epipolar Reconstruction for Assessment of Aneurysms in the Internal Carotid Artery Using In-Silico Biplane Angiograms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U7VPE6NK}},
  note         = {Machine review of arXiv:2501.11793}
}
read the original abstract

The treatment of intracranial aneurysms (IA) relies on angiography guidance using biplane views. However, accurate flow estimation and device sizing for treatment are often compromised by vessel overlap and foreshortening, which can obscure critical details. This study introduces an epipolar reconstruction approach to enhance 3D rendering of the internal carotid artery (ICA) and aneurysm dome using routinely acquired biplane angiographic data. Our method aims to improve procedural guidance by overcoming the limitations of traditional two-dimensional imaging techniques. This study employed three 3D geometries of ICA aneurysms to simulate virtual angiograms. These angiograms were generated using a computational fluid dynamics (CFD) solver, followed by the simulation of biplane angiography using a cone-beam geometry. Self-calibration was accomplished by matching contrast media position as a function of time between biplane views. Feature-matching was used to triangulate and reconstruct vascular structures in 3D. The projection data was utilized to refine the 3D estimation, including elimination of erroneous structures and ellipse-fitting. The accuracy of the reconstructions was evaluated using the Dice-Sorensen coefficient, comparing the 3D reconstructions to the original models. The proposed epipolar reconstruction method generalized well across the three tested aneurysm models, with respective Dice-Sorensen coefficients of 0.745, 0.759, and 0.654. Errors were primarily due to partial vessel overlap. The average reconstruction time for all three volumes was approximately 10 seconds. The proposed epipolar reconstruction method enhanced 3D visualization, addressing challenges such as projection-induced vessel foreshortening. This method provides a solution to the complexity of IA visualization, with the potential to provide more accurate analysis and device sizing for treatment.

Figures

Figures reproduced from arXiv: 2501.11793 by the authors.

Figure 1
Figure 1. Geometry of a biplane transmission imaging systemused for epipolar reconstruction. Unprimed coordinates are relative to the first imaging system including observer, O (the x-ray source), image plane, D, and projection p of a point P in the object, while primed coordinates are relative to the second imaging system including O’, D’ and p’ [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Time points t1 and t2 are compared in both frontal and lateral views after pixel pitch equalization. The advancement of contrast in each view is measured by taking the difference (in pixels) between the top-most location of contrast media in t2 and t1, denoted 𝑙 and 𝑙 ′ for the frontal and lateral measurements, respectively. The ratio of these values is used to equalize the magnification factor between frontal and l… view at source ↗
Figure 3
Figure 3. After magnification factor equalization, representation of contrast advancement (in pixels) should be equivalent between views. For a given time step, any remaining discrepancies in the distance from the bottom of the image to the top￾most point of contrast media between views is attributed to FOV misalignment. This measurement is used to compare and standardize the images, ensuring the FOV of both imagers is consis… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 27 canonical work pages

  1. [1]

    Gounis, Ajay K

    Manik Mehra GS, Matthew J. Gounis, Ajay K. Wakhloo. Intracranial Aneurysms: Clinical Assessment and Treatment Options. In: McGloughlin T, ed. Biomechanics and Mechanobiology of Aneurysms. Studies in Mechanobiology, Tissue Engineering and Biomaterials. Springer, Berlin, Heidelberg; 2011:331-372

    work page 2011

  2. [2]

    Vega-Quesada, Gonzalo Emiliano Aranda-Abreu Atlas of Nervous System Vascular Malformations: A Systematic Review

    Carlos Castillo-Rangel GM, Karla Aketzalli Hernandex-Contreras, Cristofer Zarate- Calderon, Micheel Merari Vichi-Ramirez, Wilmar Cortez-Saldias, Marco Antonio Rodriguez-Florido, Ambar Elizabeth Riley-Moguel, Omar Pichardo, Osvaldo Torres- Pineda, Helena G. Vega-Quesada, Gonzalo Emiliano Aranda-Abreu Atlas of Nervous System Vascular Malformations: A System...

    work page doi:10.3390/life12081199 2022

  3. [3]

    Bednarek

    Anant Gopal KRH, Stephen Rudin, Daniel R. Bednarek. Reconstruction of asymmetric vessel lumen from two views. Paper/Poster presented at: SPIE: Medical Imaging; May 9, 2002; San Diego, California

    work page 2002

  4. [4]

    Hoffmann AMW, Peter B

    Kenneth R. Hoffmann AMW, Peter B. Noel. 3D reconstruction of the carotid artery from two views using a single centerline. International Congress Series. 2004;1268:177-182. doi: https://doi.org/10.1016/j.ics.2004.03.145

    work page doi:10.1016/j.ics.2004.03.145 2004

  5. [5]

    3D Reconstruction with Coronary Artery Based on Curve Descriptor and Projection Geometry-Constrained Vasculature Matching

    Tong J, Xu S, Wang F, Qi P. 3D Reconstruction with Coronary Artery Based on Curve Descriptor and Projection Geometry-Constrained Vasculature Matching. Information. 2022;13:38. doi: 10.3390/info13010038

    work page doi:10.3390/info13010038 2022

  6. [6]

    Using temporal and structural data to reconstruct 3D cerebral vasculature from a pair of 2D digital subtraction angiography sequences

    Frisken S, Haouchine N, Du R, Golby AJ. Using temporal and structural data to reconstruct 3D cerebral vasculature from a pair of 2D digital subtraction angiography sequences. Computerized Medical Imaging and Graphics. 2022;99:102076. doi: https://doi.org/10.1016/j.compmedimag.2022.102076

    arXiv 2022

  7. [7]

    3D reconstruction of the cerebral arterial network from stereotactic DSA

    Coste E, Vasseur C, Rousseau J. 3D reconstruction of the cerebral arterial network from stereotactic DSA. Medical Physics. 1999;26:1783-1793. doi: 10.1118/1.598682 14

    work page doi:10.1118/1.598682 1999

  8. [8]

    4D interventional device reconstruction from biplane fluoroscopy

    Wagner M, Schafer S, Strother C, Mistretta C. 4D interventional device reconstruction from biplane fluoroscopy. Medical Physics. 2016;43:1324-1334. doi: 10.1118/1.4941950

    work page doi:10.1118/1.4941950 2016

Show all 30 references
  1. [9]

    A computer algorithm for reconstructing a scene from two projections

    Longuet-Higgins HC. A computer algorithm for reconstructing a scene from two projections. Nature. 1981;293:133-135. doi: https://doi.org/10.1038/293133a0

    1981 doi

  2. [10]

    Least-Squares Estimation of Transformation Parameters Between Two Point Patterns

    Umeyama S. Least-Squares Estimation of Transformation Parameters Between Two Point Patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1991;13:376-380. doi: https://doi.org/10.1109/34.88573

    1991 doi

  3. [11]

    Calibrating a Binocular Stereo Through Projective Reconstruction Using Both a Calibration Object and the Environment

    Zhengyou Zhang OF, Rachid Deriche. Calibrating a Binocular Stereo Through Projective Reconstruction Using Both a Calibration Object and the Environment. Europe-China Workshop on Geometrical modeling and Invariants for Computer Vision. 1995

    1995

  4. [12]

    Angiographic velocimetry analysis using contrast dilution gradient method with a 1000 frames per second photon-counting detector

    Williams KA, Shields A, Setlur Nagesh SV, Chudzik M, Bednarek DR, Rudin S, Ionita C. Angiographic velocimetry analysis using contrast dilution gradient method with a 1000 frames per second photon-counting detector. Journal of Medical Imaging. 2023;10. doi: 10.1117/1.jmi.10.3.033502

    2023 doi

  5. [13]

    Passive Scalars in Turbulent Flows

    Warhaft Z. Passive Scalars in Turbulent Flows. Annual Review of Fluid Mechanics. 2000;32:203-240. doi: https://doi.org/10.1146/annurev.fluid.32.1.203

    2000 doi

  6. [14]

    On Feature Tracking in X-Ray Images

    Klüppel M, Wang J, Bernecker D, Fischer P, Hornegger J. On Feature Tracking in X-Ray Images. In: Deserno TM, Handels H, Meinzer H-P, Tolxdorff T, eds. Bildverarbeitung für die Medizin 2014: Algorithmen - Systeme - Anwendungen Proceedings des Workshops vom 16. bis 18. März 2014...

    2014

  7. [15]

    The OpenCV Library

    Bradski G. The OpenCV Library. Dr Dobb's Journal of Software Tools. 2000

    2000

  8. [16]

    Smooth Vasculature Reconstruction with Circular and Elliptic Cross Sections

    Karl Krissian XW, Vincent Luboz. Smooth Vasculature Reconstruction with Circular and Elliptic Cross Sections. In: Westwood JD, ed. Medicine Meets Virtual Reality 14. Amsterdam, Netherlands: IOS Press; 2006:273-278

    2006

  9. [17]

    A 2D driven 3D vessel segmentation algorithm for 3D digital subtraction angiography data

    Spiegel M, Redel T, Struffert T, Hornegger J, Doerfler A. A 2D driven 3D vessel segmentation algorithm for 3D digital subtraction angiography data. Physics in Medicine and Biology. 2011;56:6401-6419. doi: 10.1088/0031-9155/56/19/015

    2011 doi

  10. [18]

    Characterization of velocity patterns produced by pulsatile and constant flows using 1000 fps high-speed angiography (HSA)

    Shields A, Setlur Nagesh SV, Ionita C, Bednarek DR, Rudin S. Characterization of velocity patterns produced by pulsatile and constant flows using 1000 fps high-speed angiography (HSA). SPIE Medical Imaging. 2021. doi: 10.1117/12.2580888.short

    2021 doi

  11. [19]

    Evaluation of methods to derive blood flow velocity from 1000 fps high-speed angiographic sequences (HSA) using optical flow (OF) and computational fluid dynamics (CFD)

    Shields A, Setlur Nagesh SV, Ionita CN, Bednarek DR, Rudin S. Evaluation of methods to derive blood flow velocity from 1000 fps high-speed angiographic sequences (HSA) using optical flow (OF) and computational fluid dynamics (CFD). Paper/Poster presented 2021, https://dx.doi.o...

    2021 doi

  12. [20]

    Quantifying the effects of stenosis on blood flow using 1000 fps High-Speed Angiography (HSA)

    Krebs JM, Shields A, Setlur Nagesh SV, Shepard LM, Ionita CN, Bednarek D, Rudin S. Quantifying the effects of stenosis on blood flow using 1000 fps High-Speed Angiography (HSA). Paper/Poster presented at: SPIE Medical Imaging; 2020, 2020; Houston, Texas. https://dx.doi.org/10....

    2020 doi

  13. [21]

    Application of 1,000 fps High-Speed Angiography to In-Vitro Hemodynamic Evaluation of Left Ventricular Assist Device Outflow Graft Configurations

    Shields A, Setlur Nagesh SV, Rajagopal K, Bednarek DR, Rudin S, Chivukula VK. Application of 1,000 fps High-Speed Angiography to In-Vitro Hemodynamic Evaluation of Left Ventricular Assist Device Outflow Graft Configurations. ASAIO Journal. 2023;69

    2023

  14. [22]

    First in-vivo demonstration of 1000fps High 15 Speed Coronary Angiography (HSCA) in a swine animal model

    Setlur Nagesh SV, Vanderbilt E, Koenigsknecht C, Pionessa D, Chivukula VK, Ionita CN, Zlotnick DM, Bednarek DR, Rudin S. First in-vivo demonstration of 1000fps High 15 Speed Coronary Angiography (HSCA) in a swine animal model. Paper/Poster presented 2024, https://dx.doi.org/10...

    2024 doi

  15. [23]

    Demonstration of 1000 fps high-speed angiography (HSA) in pre-clinical in-vivo rabbit aneurysm models during flow-diverter treatment

    Vanderbilt EA, Koenigsknecht C, Pionessa D, Ionita CN, Bednarek DR, Rudin S, Nagesh SVS. Demonstration of 1000 fps high-speed angiography (HSA) in pre-clinical in-vivo rabbit aneurysm models during flow-diverter treatment. Paper/Poster presented 2024, https://dx.doi.org/10.111...

    2024 doi

  16. [24]

    Improved determination of biplane imaging geometry from two projection images and its application to three‐dimensional reconstruction of coronary arterial trees

    Chen SYJ, Metz CE. Improved determination of biplane imaging geometry from two projection images and its application to three‐dimensional reconstruction of coronary arterial trees. Medical Physics. 1997;24:633-654. doi: 10.1118/1.598129

    1997 doi

  17. [25]

    Affine transformation

    Weisstein EW. Affine transformation. https://mathworld wolfram com/. 2004

    2004

  18. [26]

    A Model-Based Consecutive Scanline Tracking Method for Extracting Vascular Networks From 2-D Digital Subtraction Angiograms

    Ping Z, Chan P, Rockett P. A Model-Based Consecutive Scanline Tracking Method for Extracting Vascular Networks From 2-D Digital Subtraction Angiograms. IEEE Transactions on Medical Imaging. 2009;28:241-249. doi: 10.1109/tmi.2008.929100

    2009

  19. [27]

    Analysis of Quantitative Angiography in Intracranial Aneurysm using Projection Foreshortening Correction and Injection Bias Removal

    Mondal P, Shields A, Bhurwani MMS, Williams KA, Nagesh SVS, Siddiqui AH, Ionita CN. Analysis of Quantitative Angiography in Intracranial Aneurysm using Projection Foreshortening Correction and Injection Bias Removal. arXiv preprint arXiv:241108185. 2024

    2024

  20. [28]

    2D versus 3D comparison of angiographic imaging biomarkers using computational fluid dynamics simulations of contrast injections

    Shields A, Shiraz Bhurwani MM, Williams K, Chivukula VK, Bednarek DR, Rudin S, Ionita C. 2D versus 3D comparison of angiographic imaging biomarkers using computational fluid dynamics simulations of contrast injections. Paper/Poster presented 2023, 2023; https://dx.doi.org/10.1...

    2023 doi

  21. [29]

    Initial evaluation of 2D and 3D simulated high-speed 1000 fps vascular contrast-flow image sequences using computational fluid dynamics (CFD)

    Shields A, Williams K, Veeturi SS, Tutino V, Ionita CN, Bednarek D, Rudin S. Initial evaluation of 2D and 3D simulated high-speed 1000 fps vascular contrast-flow image sequences using computational fluid dynamics (CFD). Paper/Poster presented 2022, 2022; https://dx.doi.org/10....

    2022 doi

  22. [30]

    Use of high-speed angiography HSA-derived boundary conditions and Physics Informed Neural Networks (PINNs) for comprehensive estimation of neurovascular hemodynamics

    Williams KA, Shields A, Shiraz Bhurwani MM, Nagesh SVS, Bednarek DR, Rudin S, Ionita CN. Use of high-speed angiography HSA-derived boundary conditions and Physics Informed Neural Networks (PINNs) for comprehensive estimation of neurovascular hemodynamics. Paper/Poster presente...

    2023 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.