Pith. sign in

REVIEW 4 major objections 7 minor 23 references

Enhancing the automatic segmentation and analysis of 3D liver vasculature models

T0 review · 4 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read An automatic pipeline using a learned centerline network and connected-component heuristics separates liver CT vessel segmentations into surgeon-validated portal and hepatic trees, and releases a 77-case annotated dataset enabling…

desk verdict Useful pipeline and honest negative result, but the dataset count doesn't add up and the 'clean low-error' claim only covers the cases the algorithm handled. read the letter →

arxiv 2411.15778 v4 pith:X4SNCC5T submitted 2024-11-24 eess.IV cs.AIcs.CV

Yassine Machta , Omar Ali , Kevin Hakkakian , Ana Vlasceanu , Amaury Facque , Nicolas Golse , Irene Vignon-Clementel This is my paper · ORCID
classification eess.IVcs.AIcs.CV
keywords livervesselsegmentationportalveinhepaticskeletonizationClDiceconnectedcomponentsmorphometrymedicalimagingdataset
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

The paper tries to establish that an end-to-end pipeline—a pretrained neural skeletonizer used inside a topology-preserving loss, followed by a connected-component and skeleton-based separation algorithm—can turn a single-class 3D segmentation of liver vessels into anatomically labeled portal and hepatic venous trees with low error. The authors show that their neural skeletonizer represents vessel centerlines much better than morphological thinning, with 98.2% surface Dice versus 86.0% and 81.8% for the morphological baselines, and they build a public dataset of 77 CT cases with separated and branch-labeled trees. A sympathetic reading of the results is that the bottleneck for accurate liver vessel models is not skeleton quality within the ClDice loss, since adding NeuralSkel there does not improve segmentation, but rather having a large multi-class training set, which their separation algorithm provides. This matters because surgery planning and disease assessment in liver cancer depend on knowing which vessel perfuses and which drains each part of the liver, and such annotated data are scarce.

What carries the argument

Two mechanisms carry the argument. NeuralSkel is a pretrained centerline-prediction network built on the nnU-Net architecture; it is differentiable, so it can replace morphological skeletonization inside the ClDice loss, and it produces centerlines that stay close to the reference skeleton, reaching 98.2% surface Dice versus 86.0% for soft-skeletonization and 81.8% for soft-persistent-skeletonization. The separation algorithm is the second mechanism: it pastes an inferred two-class labeling onto the original segmentation, dilates it, isolates uncovered regions as connected components, assigns them to the nearest tree, skeletonizes components touching both trees, splits them into branches, and resolves each branch by an angle rule of less than 60 degrees, iterating until no conflict remains.

What would settle it

Take the nine excluded LiTS cases, repair the single-class segmentation so the distal regions are connected, rerun the separation algorithm, and count voxels that surgeons still judge mislabeled; if the error remains concentrated in the dense distal periphery, the angle-and-proximity heuristic is the bottleneck rather than the initial segmentation.

Watch

Extended reading notes

Core claim

The central claim is that a learnable skeletonizer, NeuralSkel, a pretrained nnU-Net that predicts centerlines from softmax segmentation volumes, gives markedly better centerline geometry than the morphological skeletonization used by prior differentiable losses, but the paper reports that plugging this better skeleton into the ClDice loss does not improve the final vessel segmentation. Instead, the discovery that carries the contribution is algorithmic: after pasting a model's inferred portal and hepatic labeling onto a single-class annotation, disputed regions are resolved by connected-component proximity and a 60-degree branch-angle criterion derived from Murray's law, yielding a two-tree separation that surgeons rated as flawless or mildly confused in 53 of 62 LiTS cases. On the resulting LIRCAD dataset, retraining the vessel segmentation model improves all metrics by two to three percentage points and reduces Hausdorff distance. The paper further claims that these separated trees can be automatically labeled into clinically named branches, including the main portal vein, right and left branches, and Couinaud segments, and then analyzed morphometrically.

Load-bearing premise

The load-bearing premise is that contested voxels can be assigned to the correct venous tree from connected-component proximity and a fixed 60-degree branch-angle rule; if dense, highly interconnected distal regions defeat that premise, as the paper itself reports for 9 of 62 LiTS cases, the claimed clean multi-class segmentation is only established for the cases that survive such heuristics.

Editorial extensions

If this is right

  • If NeuralSkel generalizes as a skeletonizer, any tubular-structure segmentation trained with ClDice can swap in a learned, more accurate centerline module without hand-tuning morphological kernels.
  • The LIRCAD dataset of 77 separated, branch-labeled CT cases gives the community training data that improves vessel segmentation by two to three percentage points over training on IRCAD alone.
  • Automatic anatomical branch labels make morphometric measurements such as length, radius, emergence angle, generation, and descendant counts available per branch, enabling quantitative comparisons of right versus left liver territories.
  • The separation algorithm reduces annotation burden: surgeons needed to fix only minor mix-ups in 18 of the 53 final cases rather than labeling tree membership from scratch.
  • Because the separation works from a single-class input, it can be reapplied to other single-class annotations of liver vessels, not just the in-house LiTS subset.

Reading between the lines

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

  • The paper's own exclusion of nine LiTS cases suggests the separation heuristics are reliable only where the initial segmentation is clean; a testable extension is to run the algorithm on a segmentation with better distal connectivity and check whether those nine cases become resolvable.
  • Since skeleton quality improved but downstream ClDice segmentation did not, the topology signal that matters may already be captured by the simpler losses; pairing NeuralSkel with other topology-preserving losses or using it to supervise post-processing could yet translate its centerline accuracy into segmentation gains.
  • The 60-degree branch-angle rule is organ-agnostic in principle, so the same separation tool could be tried on other paired vascular trees, such as hepatic artery and bile ducts, where anatomy follows similar branching, with manual validation.
  • The morphometric features extracted here could be correlated with tumor location and liver regeneration outcomes in future work, but that clinical link is not established by this paper.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper presents an end-to-end pipeline for 3D liver vessel analysis: an nnU-Net-based skeletonization network (NeuralSkel) trained to predict Lee's skeletonization and used in a ClDice-style loss; an algorithm that separates single-label vessel segmentations into portal and hepatic trees using connected components, skeleton analysis, and a 60-degree branch-angle heuristic; automatic anatomical branch labeling; and a new public dataset called LIRCAD. The results show that NeuralSkel approximates Lee's skeleton more closely than morphological soft-skeletonization methods, but that using it inside the ClDice loss does not improve overall segmentation performance. Training on the automatically separated LIRCAD labels improves segmentation metrics by 2-3 points over the IRCAD-only baseline. The paper also reports surgeon-based validation of the separation and a morphometric analysis of the labeled trees.

Significance. If the claims hold, the paper would make three useful contributions: a reproducible skeletonization network for tubular structures, a practical tool for converting single-class vessel annotations into multi-class portal/hepatic labels, and a public dataset with anatomical branch labels and morphometric features. The explicit release of code and data and the surgeon-in-the-loop validation are strengths, as is the transparency about the negative ClDice result. However, the significance is tempered by two load-bearing issues: the skeletonization comparison is partly circular because NeuralSkel is trained on the same Lee skeletonization used as the evaluation reference, and the 'clean, low-error' dataset claim is demonstrated only on a selected subset, since 9 of 62 LiTS cases were excluded and the dataset count is internally inconsistent (77 claimed vs 73 derivable from the paper's own counts).

major comments (4)
  1. [Abstract and §3.2/§3.3] The abstract and §3.3 state that the new public dataset contains 77 cases, but §3.2 reports 53 LiTS volumes after validation plus 20 IRCAD volumes, which sum to 73. This discrepancy is load-bearing because the dataset size is a headline contribution. Please reconcile the numbers, explain the source of the additional 4 cases, or correct the count to 73.
  2. [Abstract and §3.2] The abstract's claim of a 'clean multi-class segmentation of the vessels, validated by surgeons to provide low error' is supported only for the 53 retained LiTS cases: 9 of 62 cases were explicitly excluded because they presented 'larger errors' in dense interconnected regions. No quantitative error metric is reported for the excluded cases or for the full 62-case set, so the claim overstates the algorithm's generality, and the training improvement in Table 2 may be partly attributable to selection. Please evaluate the separation algorithm on all 62 cases, report per-case failure rates, or revise the claim to describe the subset where the heuristics hold.
  3. [§2.1 and Table 1] NeuralSkel is trained to predict Lee's skeletonization, and Table 1 compares all skeletonization methods against Lee's output. Consequently, the reported 98.2% Surface Dice for NeuralSkel is a fidelity measure to the training target, not an independent assessment of centerline quality. The Discussion's statement that NeuralSkel gives 'much better centerline representations than the morphological operations' should be rephrased as 'closer to Lee's skeleton' unless an independent reference (e.g., manually traced centerlines or a different skeletonization algorithm) is used.
  4. [§2.2 and §3.2] The separation algorithm relies on a fixed 60-degree branch-angle threshold and connected-component proximity (step 9, citing Murray's law), and the Discussion acknowledges that these fail in dense distal regions. The validation consists of a surgeon's categorical classification ('flawless', 'small mixup', 'larger errors') rather than a quantitative per-voxel or per-tree error. Please provide quantitative metrics (e.g., assignment accuracy, volume overlap against a manual reference) on a per-case basis, including the excluded cases, so that the 'low error' claim is measurable and not just categorical.
minor comments (7)
  1. [§2.2] The term 'Bersenham line' appears to be a typo for 'Bresenham line'.
  2. [Code and data availability] The manuscript contains placeholder links 'Code link' and 'Data link'; the final version should provide actual URLs and a clear data availability statement.
  3. [§2.3] The automatic anatomical labeling heuristics, such as fusing paths based on Couinaud segments, are described only briefly; a flowchart or pseudo-code would improve reproducibility.
  4. [Figure 2] The step numbers in the text (2)-(12) do not clearly correspond to the numbered boxes in the figure; please align the numbering between text and figure.
  5. [§3.1] The Surface Dice tolerance is fixed at 2 mm; please justify this value or state it explicitly as a tunable parameter.
  6. [§2.1] Training details for NeuralSkel (dataset, number of volumes, augmentation strategy, number of epochs) are not specified; adding these details would improve reproducibility.
  7. [§3.3] The claim of the 'largest public liver vessel dataset to date' should be substantiated by comparing with existing public datasets, such as LiTS vessel annotations, IRCAD, or other published vessel datasets.

Circularity Check

1 steps flagged · score 3.0 of 10

Minor circularity in the NeuralSkel evaluation (trained on Lee, measured on Lee); the central separation and dataset claims rest on independent surgeon validation.

  1. fitted input called prediction [Section 2.1; Table 1 caption; Discussion §4]
    "an nnunet [18] was pretrained to predict the liver vessel centerlines ... The ground truth (GT) leveraged for the training of the network is the centerline of the different liver vessels obtained with Lee's [17] skeletonization method ... Table 1: Comparison of skeletonization methods with respect to the output of Lee's alogrithm."

    NeuralSkel's training labels are Lee centerlines, and Table 1 evaluates methods explicitly 'with respect to the output of Lee's algorithm.' The reported gains (SurfaceDice 98.2% vs 86.0/81.8; Dice 34.3 vs 23.59/20.15) therefore measure how well a network optimized to imitate Lee reproduces Lee on held-out volumes, not an independent notion of centerline quality. The morphological baselines are not trained against Lee, so the comparison is by construction favorable to the learned method. The Discussion's 'much better centerline representations' overstates this as a general property; it is a fit-to-target evaluation. This is partial circularity, not a full reduction of the paper's central claims.

full rationale

No load-bearing self-citation chain or imported uniqueness theorem is present. The separation algorithm's 'low error' claim is anchored to independent surgeon review of the output (35 flawless, 18 minor mixups fixed into 53 LiTS volumes), which is an external benchmark. The branch labeling and morphometry rely on external libraries and the Tian et al. Couinaud dataset. The main circularity is confined to the NeuralSkel comparison in Table 1, which contrasts a Lee-trained network against non-Lee-trained morphological skeletonizers using Lee-derived ground truth; this inflates the 'better centerline' claim but does not determine the dataset or separation results. The 9 discarded LiTS cases and the 77-case count (53+20=73) are internal-consistency and correctness issues, not circularity. Overall score 3.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claims depend on several hand-selected thresholds and unproven domain assumptions about liver vascular anatomy, plus the transferability of an external segmenter; no code, data, or trained models are available to check these choices. No new physical entities are introduced.

free parameters (6)
  • branch connection angle threshold = 60 degrees
    Branches are considered connected when the angle between them is below 60 degrees, a fixed threshold taken from Murray's law [14] and applied to venous trees in voxel space (Methods, §2.2, step 9). The result of the separation algorithm depends on it.
  • minimum connected component size = 50 voxels
    Connected components smaller than 50 voxels are removed as floating patches before tree assignment (Methods, §2.2, step 5). The value is chosen by hand.
  • distance cutoff for outlier classes = unspecified
    Components farther than both trees by a fixed distance are filtered to avoid floating patches (Methods, §2.2); the distance is not reported, so the criterion is not reproducible.
  • dilation of the inferred dual-label segmentation = unspecified
    The inferred segmentation is dilated to compensate for undersegmentation before being pasted onto the original single-label volume (Methods, §2.2); the kernel size is not stated.
  • Surface Dice tolerance = 2 mm
    Surface Dice is computed with a tolerable threshold fixed at 2 mm (§3.1). This evaluation choice affects all reported Surface Dice numbers.
  • skeleton branch pruning criteria = heuristic, unspecified
    Extremely short paths and loops produced by skeletonization and sknw are fused or removed with unspecified rules (§2.3, §3.3); these pruning decisions affect the morphometry graph and measurements.
assumptions (5)
  • domain assumption Lee's skeletonization provides valid ground-truth centerlines for liver vessel trees
    Lee's algorithm [17] is used to generate the training target for NeuralSkel and to evaluate it in Table 1 (§2.1, §3.1). If the reference skeleton is anatomically wrong, both the learned model and the benchmark inherit the error.
  • domain assumption Portal and hepatic venous trees are mostly disconnected in CT, with only local conflicts
    The separation algorithm isolates conflict regions as connected components and assigns them to the closest tree (§2.2). Dense interconnections caused failures in 9 of 62 LiTS cases (§3.2), so this assumption does not always hold.
  • domain assumption Murray's law branching-angle rule (60 degrees) applies to liver venous trees at CT resolution
    A single 60-degree threshold, motivated by arterial branching physiology [14], is used to decide whether branches belong to the same tree (§2.2, step 9). Venous hemodynamics and voxel discretization may invalidate it.
  • domain assumption Vessel tree roots are the endpoints farthest from the liver segmentation, i.e., outside the liver
    Tree roots are automatically detected as the farthest tree endpoints from the liver volume (§2.3). For the portal vein at the hilum, the root may lie inside the liver mask, making the assumption case-dependent.
  • domain assumption The external GLC-UNet Couinaud segmenter (Tian et al. [16]) generalizes to LIRCAD CT volumes
    Couinaud segment predictions from a model trained on another dataset are used to assign anatomical names to branches (§2.3). No accuracy of this transfer is reported for the LIRCAD cases.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing the automatic segmentation and analysis of 3D liver vasculature models." pith.science (2026). https://pith.science/paper/X4SNCC5T

@misc{pith2026241115778,
  author       = {Pith},
  title        = {Pith review of: Enhancing the automatic segmentation and analysis of 3D liver vasculature models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X4SNCC5T}},
  note         = {Machine review of arXiv:2411.15778}
}
read the original abstract

Surgical assessment of liver cancer patients requires identification of the vessel trees from medical images. Specifically, the venous trees - the portal (perfusing) and the hepatic (draining) trees are important for understanding the liver anatomy and disease state, and perform surgery planning. This research aims to improve the 3D segmentation, skeletonization, and subsequent analysis of vessel trees, by creating an automatic pipeline based on deep learning and image processing techniques. The first part of this work explores the impact of differentiable skeletonization methods such as ClDice and morphological skeletonization loss, on the overall liver vessel segmentation performance. To this aim, it studies how to improve vessel tree connectivity. The second part of this study converts a single class vessel segmentation into multi-class ones, separating the two venous trees. It builds on the previous two-class vessel segmentation model, which vessel tree outputs might be entangled, and on connected components and skeleton analyses of the trees. After providing sub-labeling of the specific anatomical branches of each venous tree, these algorithms also enable a morphometric analysis of the vessel trees by extracting various geometrical markers. In conclusion, we propose a method that successfully improves current skeletonization methods, for extensive vascular trees that contain vessels of different calibers. The separation algorithm creates a clean multi-class segmentation of the vessels, validated by surgeons to provide low error. A new, publicly shared high-quality liver vessel dataset of 77 cases is thus created. Finally a method to annotate vessel trees according to anatomy is provided, enabling a unique liver vessel morphometry analysis.

Figures

Figures reproduced from arXiv: 2411.15778 by the authors.

Figure 1
Figure 1. Model description for multi-class vessel segmentation (right box) which [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Vascular tree separation algorithm; CC: Connected component. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Vessel tree skeletonization methods (a) Lee’s skeletonization (b) Soft-skel (c) Soft-persistent-skel (d) Neural skel 3 Results 3.1 DL techniques for tree skeletonization and segmentation Both models’(skeletonization and segmentation) results are evaluated with the Dice, ClDice, Surface Dice [22] and Hausdorff distance metrics, after taking into account the CT sampling. The tolerable threshold for the surface Dice is… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Annotation. Visualizations by Slicer 3D (www.slicer.org) [23] [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Morphometry - portal tree properties  [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 2
Figure 2. Figure 2: Fig.2. Since it occurs mostly in the distal parts of the tree, its impact for clinical [PITH_FULL_IMAGE:figures/full_fig_p008_2.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 16 canonical work pages

  1. [1]

    Ali et al., CoRe: An Automated Pipeline for the Prediction of Liver Resec- tion Complexity from Preoperative CT Scans, In: S.T

    O. Ali et al., CoRe: An Automated Pipeline for the Prediction of Liver Resec- tion Complexity from Preoperative CT Scans, In: S.T. Kakileti et al., eds.,Arti- ficial Intelligence over Infrared Images for Medical Applications and Medical Im- age Assisted Biomarker Discovery. MIABID AIIIMA 2022 2022, Lecture Notes in Computer Science, vol. 13602, Springer, ...

    work page 2022

  2. [2]

    Rumgay, et al.,Global burden of primary liver cancer in 2020 and predictions to 2040, Journal of Hepatology, vol

    H. Rumgay, et al.,Global burden of primary liver cancer in 2020 and predictions to 2040, Journal of Hepatology, vol. 77, no. 6, pp. 1598–1606, Dec. 2022, doi:10.1016/ j.jhep.2022.08.021

    work page 2020

  3. [4]

    Pro- moting Connectivity of Network-Like Structures by Enforcing Region Separation,

    D. Oner, M. Koziński, L. Citraro, N. C. Dadap, A. G. Konings and P. Fua, "Pro- moting Connectivity of Network-Like Structures by Enforcing Region Separation," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 9, pp. 5401-5413, 1 Sept. 2022, doi: 10.1109/TPAMI.2021.3074366

    arXiv 2022

  4. [5]

    Jerman, T., Pernus, F., Likar, B., & Spiclin, Z. (2016). Enhancement of Vascular Structures in 3D and 2D Angiographic Images.IEEE Trans Med Imaging, 35(9), 2107-2118. doi: 10.1109/TMI.2016.2550102

    work page arXiv 2016

  5. [6]

    Dhruv Keshwani, Yuichiro Kitamura, Shunya Ihara, Shunta Iizuka, and Edgar Simo- Serra, TopNet: Topology Preserving Metric Learning for Vessel Tree Reconstruction and Labelling, arXiv:2009.08674 [cs], Sep. 2020. Accessed: Feb. 22, 2022. [Online]. Available: http://arxiv.org/abs/2009.08674

    work page Pith review arXiv 2009

  6. [7]

    S. Shit, J. C. Paetzold, A. Sekuboyina, I. Ezhov, A. Unger, A. Zhylka, J. PW Pluim, U. Bauer, and B. H Menze,ClDice-a Novel Topology-Preserving Loss Function for Tubular Structure Segmentation, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 16560–16569, 2021

    work page 2021

  7. [8]

    M. Viti, H. Talbot, B. Abdallah, E. Perot, and N. Gogin,Coronary Artery Cen- terline Tracking with the Morphological Skeleton Loss, in Proceedings of the IEEE International Conference on Image Processing (ICIP),pp.2741–2745,October2022. https://doi.org/10.1109/ICIP46576.2022.9897385

    arXiv

  8. [9]

    A skeletonization algorithm for gradient-based optimization

    Martin J. Menten, Johannes C. Paetzold, Veronika A. Zimmer, Suprosanna Shit, Ivan Ezhov, Robbie Holland, Monika Probst, Julia A. Schnabel, and Daniel Rueck- ert. A skeletonization algorithm for gradient-based optimization. arXiv preprint arXiv:2309.02527, 2023. https://arxiv.org/abs/2309.02527

    work page Pith review arXiv 2023

Show all 23 references
  1. [10]

    U-Net based skeletonization and bag of tricks,

    N. H. Nguyen, "U-Net based skeletonization and bag of tricks," 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW), Montreal, BC, Canada, 2021, pp. 2105-2109, doi: 10.1109/ICCVW54120.2021.00238. keywords: Deep learning;Training;Computer vision;Shape;Con...

    2021

  2. [11]

    1115–1118, 2014, doi:10.1109/ISBI.2014.6868070

    Xin Kang, Qian Zhao, Karun Sharma, Raj Shekhar, Bradford Wood, and Marius George Linguraru,Automatic labeling of liver veins in CT by probabilistic backward tracing, 2014 IEEE 11th International Symposium on Biomedical Imaging, ISBI 2014, pp. 1115–1118, 2014, doi:10.1109/ISBI....

    2014

  3. [12]

    Bilic et al.,The Liver Tumor Segmentation Benchmark (LiTS), Medical Image Analysis, Volume 84, 2023, 102680, ISSN 1361-8415,https://doi.org/10.1016/ j.media.2022.102680, 2022

    P. Bilic et al.,The Liver Tumor Segmentation Benchmark (LiTS), Medical Image Analysis, Volume 84, 2023, 102680, ISSN 1361-8415,https://doi.org/10.1016/ j.media.2022.102680, 2022

    2023

  4. [13]

    A new Python library to analyse skeleton images confirms malaria parasite remodelling of the red blood cell membrane skeleton

    JuanNunez-Iglesias,AdamJBlanch,OliverLooker,MatthewWDixon,andLeann Tilley. "A new Python library to analyse skeleton images confirms malaria parasite remodelling of the red blood cell membrane skeleton."PeerJ, vol. 6, p. e4312, Feb. 15, 2018. DOI: 10.7717/peerj.4312. PMID: 294...

    2018 doi

  5. [14]

    THE PHYSIOLOGICAL PRINCIPLE OF MINIMUM WORK AP- PLIED TO THE ANGLE OF BRANCHING OF ARTERIES

    Murray CD. THE PHYSIOLOGICAL PRINCIPLE OF MINIMUM WORK AP- PLIED TO THE ANGLE OF BRANCHING OF ARTERIES. J Gen Phys- iol. 1926 Jul 20;9(6):835-41. doi: 10.1085/jgp.9.6.835. PMID: 19872299; PMCID: PMC2140901

    1926 doi

  6. [15]

    Com- plete removal of the tumor-bearing portal territory decreases local tumor recur- rence and improves disease-specific survival of patients with hepatocellular car- cinoma

    Junichi Shindoh, Masatoshi Makuuchi, Yutaka Matsuyama, Yoshihiro Mise, Ju- nichi Arita, Yoshihiro Sakamoto, Kiyoshi Hasegawa, and Norihiro Kokudo. "Com- plete removal of the tumor-bearing portal territory decreases local tumor recur- rence and improves disease-specific surviva...

    2016 doi

  7. [16]

    Jiang Tian, Li Liu, Zhongchao Shi, and Feiyu Xu. 2019. Automatic Couinaud Segmentation from CT Volumes on Liver Using GLC-UNet. In Machine Learning in Medical Imaging: 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 13, 20...

    2019 doi

  8. [17]

    Lee, R.L

    T.C. Lee, R.L. Kashyap, C.N. Chu,Building Skeleton Models via 3-D Medial Sur- face Axis Thinning Algorithms, CVGIP: Graphical Models and Image Processing, Volume 56, Issue 6, pp. 462–478, 1994, ISSN 1049-9652,https://doi.org/10. 1006/cgip.1994.1042

    1994

  9. [18]

    Isensee, P.F

    F. Isensee, P.F. Jaeger, S.A. Kohl, J. Petersen, and K.H. Maier-Hein,nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation, Nature methods, 18(2), pp. 203–211, 2021

    2021

  10. [19]

    Soler, A

    L. Soler, A. Hostettler, V. Agnus, A. Charnoz, J. Fasquel, J. Moreau, A. Osswald, M. Bouhadjar, and J. Marescaux,3D image reconstruction for comparison of al- gorithm database: A patient specific anatomical and medical image database, Tech. Rep., IRCAD, Strasbourg, France, 2010

    2010

  11. [20]

    Méghane Decroocq, Guillaume Lavoué, Makoto Ohta, and Carole Frindel,A Soft- ware to Visualize, Edit, Model and Mesh Vascular Networks, 2022 44th Annual International Conference of the IEEE Engineering in Medicine Biology Society (EMBC), Scottish Event Campus, Glasgow, UK, July...

    2022

  12. [21]

    Strahler, A. (1957). Quantitative analysis of watershed geomor- phology. Transactions, American Geophysical Union , 38, 913-920. https://doi.org/10.1029/TR038i006p00913

    1957 doi

  13. [22]

    Ronneberger, O

    Nikolov, S., Blackwell, S., Zverovitch, A., Mendes, R., Livne, M., De Fauw, J., ... Ronneberger, O. (2021). Deep learning to achieve clinically applicable segmentation of head and neck anatomy for radiotherapy.arXiv:1809.04430 [cs.CV]

    2021 arXiv

  14. [23]

    Magnetic Resonance Imaging

    Fedorov A., Beichel R., Kalpathy-Cramer J., Finet J., Fillion-Robin J-C., Pujol S., Bauer C., Jennings D., Fennessy F., Sonka M., Buatti J., Aylward S.R., Miller J.V., PieperS.,KikinisR.3DSlicerasanImageComputingPlatformfortheQuantitative Paper accepted at MICCAI 2024 Workshop...

    2024

  15. [2023]

    https://arxiv.org/abs/2304.14572

    arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.