Multi Scale Supervised 3D U-Net for Kidney and Tumor Segmentation

Wenshuai Zhao; Zengfeng Zeng

REVIEW 4 major objections 5 minor 14 references

Multi Scale Supervised 3D U-Net for Kidney and Tumor Segmentation

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read A deliberately plain five-layer 3D U-Net, upgraded only in training details and post-processing, placed 7th of 106 teams in the KiTS19 kidney and tumor segmentation challenge.

desk verdict Solid KiTS19 challenge entry with a clear pipeline, but the central claim about multi-scale supervision is not supported by any ablation. read the letter →

arxiv 1908.03204 v2 pith:BM6CHOIT submitted 2019-08-09 eess.IV cs.CVcs.LG

Wenshuai Zhao , Zengfeng Zeng This is my paper · ORCID

classification eess.IVcs.CVcs.LG

keywords 3DU-Netkidneysegmentationtumormulti-scalesupervisionexponentiallogarithmiclossconnected-componentpost-processingKiTS19challengeCTimage

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 attempts to establish that a deliberately plain five-layer 3D U-Net, without residual, dense, attention, or feature-pyramid modules, is enough to reach near top-tier accuracy in kidney and tumor segmentation when training and post-processing are handled carefully. On the KiTS19 challenge the method placed 7th among 106 teams with a composite Dice of 0.8961, kidney Dice 0.9741, and tumor Dice 0.8181; on the paper's own 42-patient split the corresponding scores are 0.969 for kidney and 0.805 for tumor. The contribution is that three practical ingredients make the plain network work: multi-scale supervision in the decoder, an exponential logarithmic loss weighted toward the hard tumor class, and a connected-component post-processing step that removes kidney or tumor regions not attached to a kidney. This matters because novel architecture blocks often overfit to the dataset where they were tuned, while a stable base architecture with thoughtful training and task-specific priors gives a more transferable recipe.

What carries the argument

The load-bearing mechanism is multi-scale supervision in the decoder of a five-layer 3D U-Net: each decoder stage emits a segmentation map at its own resolution and the map is compared with the label downsampled to that resolution, so the auxiliary losses push every decoding level to be correct rather than letting the final layer compensate for coarse errors. Strided convolutions replace pooling for downsampling and transposed convolutions replace trilinear interpolation for upsampling, with the base feature count set to 30. Around this sit the exponential logarithmic loss, combining soft Dice and cross-entropy with per-class weights ($0.4$ for kidney and $0.6$ for tumor in the Dice term; $0.28$, $0.28$, $0.44$ for background, kidney, and tumor in the cross-entropy term), and the connected-component post-processing that keeps only the largest kidney components and any tumor attached to them.

What would settle it

On the paper's 42-patient validation set, rerun the final model with the auxiliary multi-scale decoder losses removed and with the connected-component post-processing removed, and compare kidney and tumor Dice with the reported 0.969 and 0.805; also list ground-truth tumors that are not connected to the ground-truth kidney. If scores do not drop without the decoder supervision, the proposed mechanism is not the source of the gain, and if any disconnected ground-truth tumor is deleted by post-processing, the anatomical prior is violated and the reported tumor score is optimistic for those cases.

Watch

Extended reading notes

Core claim

The paper's central claim is that a five-layer 3D U-Net trained with three practical enhancements is a top-tier kidney and tumor segmentation method. The enhancements are: supervising every decoder resolution with a downsampled label so the network cannot postpone correct prediction to the final upsampling; using an exponential logarithmic loss with manually chosen class weights that up-weight the small, morphologically variable tumor class; and applying a connected-component post-processing rule that keeps only the largest kidney-connected components and removes scattered voxels. The supporting evidence is the KiTS19 challenge result, Composite Dice 0.8961 with kidney 0.9741 and tumor 0.8181, ranking 7th of 106 teams, together with the local 42-patient split scores of 0.969 for kidney and 0.805 for tumor, with low variance for kidney. The paper also argues that medical images are less diverse than natural images, so the shallow plain U-Net does not need the architectural complexity common in natural-image segmentation.

Load-bearing premise

The reported scores rely on the assumption that every true kidney region belongs to one of at most two kidney components and that every true tumor voxel lies inside or attached to a kidney component; when that fails, the connected-component post-processing deletes correct voxels and the reported tumor Dice is no longer representative of those patients.

Editorial extensions

If this is right

Near top-tier kidney and tumor segmentation can be reached with a five-layer plain 3D U-Net, so the main engineering effort shifts from architecture invention to loss design, augmentation, and dataset-specific post-processing.
Multi-scale decoder supervision extracts more learning signal from a limited labeled set by supervising every resolution instead of only the final one.
The exponential logarithmic loss with manual class weights gives a practical recipe for volumetric tasks in which the target structure is rare, as tumor is relative to kidney and background.
Post-processing built on anatomical priors is a high-value, low-cost step, and its benefit is tied to how often the prior holds in the test population.
The remaining performance gap in the challenge is concentrated in tumor segmentation, not kidney segmentation, pointing to tumor morphology as the main next target.

Reading between the lines

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

A direct test the paper leaves implicit is an ablation that removes the multi-scale decoder losses while keeping the loss weighting and post-processing; if the paper's thesis is right, composite Dice should fall, with the largest drop on small and morphologically variable tumors.
The same recipe should transfer to other volumetric segmentation problems with small annotated sets and unbalanced target structures, such as pancreas or lymph-node segmentation; the paper does not claim this transfer.
The connected-component prior can be turned into a quantitative failure analysis: counting how often ground-truth tumor is anatomically separate from the kidney in the challenge data sets an upper bound on how much the post-processing can help and how much error it can introduce.
If the paper's premise that medical data need shallow networks is right, adding depth or dense connections to comparable tasks should yield diminishing returns under the same training protocol; this is a conjecture beyond the paper's evidence.

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.

Desk Editor's Note

read the letter

Dear colleague,

This is a short KiTS19 challenge report, not a methods paper. The concrete empirical result — a 5-layer 3D U-Net with multi-scale decoder supervision, exponential logarithmic loss, and connected-component postprocessing, ranked 7th of 106 teams with composite Dice 0.8961 (kidney 0.9741, tumor 0.8181) — is honestly reported and internally consistent. The local split scores (0.969 and 0.805) sit plausibly with the challenge scores, and the paper does not misstate what the leaderboard says.

What the paper does well is describe the full pipeline: intensity clipping, resampling to isotropic spacing, patch sampling, augmentation, training schedule, mirrored inference, and the postprocessing rule. The writing is clear and the component choices are grounded in the literature (nnU-Net, exponential logarithmic loss, etc.). I found no numerical errors.

The soft spot is attribution, and this is where the stress-test concern lands. The title and Section 2.2 claim multi-scale supervision 'encourages the network to predict correctly from deep layers' and improves the final prediction, but no ablation isolates that component. All three enhancements are applied as a single bundle, so the reported Dice cannot be assigned to the supervision scheme. The gain could come from the exponential-log loss, the class weights, the augmentation, the postprocessing, or the particular 42-patient split. There are also no error bars, no repeated-seed variance, and no code or artifacts shipped; the split is described only as 'we divided 42,' with no stratification details. The connected-component postprocessing assumes at most two kidneys and that tumor is attached to kidney; if a true disconnected tumor fragment exists, those true-positive voxels get removed, and the paper does not analyze how often that happens. These are real but proportionate weaknesses: the benchmark number itself is externally measured, and the causal claim is the only load-bearing part that fails.

I would want ablations and error bars before taking the causal claim seriously. But for a challenge report, the empirical result stands. I would send this to a serious referee rather than desk reject, with a request for at least one baseline without multi-scale supervision and a repeated-seed variance estimate. If the supervision helps, the paper becomes a useful reproducibility data point; if not, it is still a clean description of a competitive single model.

Who is this for? Anyone doing kidney/tumor segmentation or looking for a solid 3D U-Net baseline. I would not cite it in my own work, but it is worth a reading-group slot as a compact example of a transparent challenge entry.

Hope this helps.

Referee Report

4 major / 5 minor

Summary. The paper proposes a multi-scale supervised 3D U-Net for kidney and tumor segmentation in the KiTS19 challenge. The architecture is a five-layer 3D U-Net in which decoder outputs at multiple resolutions are supervised by corresponding resolution labels, trained with an exponential logarithmic Dice-plus-cross-entropy loss, followed by a connected-component postprocessing step that removes kidney/tumor detections not attached to the main kidney. In a local split of 42 of the 210 public training patients, the method reports Dice coefficients of 0.969 (kidney) and 0.805 (tumor); on the challenge test set it reports a composite Dice of 0.8961 (0.9741 kidney, 0.8181 tumor), ranking 7th among 106 teams. The paper attributes the performance mainly to multi-scale supervision, class-imbalance handling, and postprocessing, but provides no ablation isolating these contributions.

Significance. If the reported challenge result is correct, the paper shows that a relatively simple, carefully trained 3D U-Net without architectural bells and whistles can be highly competitive on a challenging medical segmentation benchmark. The reported local Dice values and the challenge composite Dice are internally consistent: 0.8961 is the mean of 0.9741 and 0.8181, and the kidney variance is stated to be small. The main scientific contribution claimed, namely that multi-scale decoder supervision improves the final full-resolution prediction, is plausible and of interest to the medical image segmentation community. However, as presented, that claim is not supported by any control experiment, and the lack of an ablation limits the paper's contribution to a single competitive result rather than an evidence-based methodological finding. The paper does not provide code or detailed experimental protocols that would allow independent verification or reuse.

major comments (4)

[Abstract and Section 2.2 (Network Architecture)] The central methodological claim—that multi-scale supervision 'encourages the network to predict right results from the deep layers' and 'improved the performance in the final layer'—is not supported by any ablation or control experiment. Section 3 reports only the full configuration with multi-scale supervision, exponential logarithmic loss, and postprocessing applied together, and no baseline 3D U-Net without the auxiliary decoder losses is trained, evaluated, or reported. Consequently, the reported Dice scores cannot be attributed to multi-scale supervision; they could be driven by the loss formulation, the class weighting, the data augmentation, the postprocessing, or the particular 42-patient split. This is a load-bearing issue because the paper's stated contribution is the supervision scheme, not the benchmark number alone.
[Section 3 (Experiments and Results)] The internal test split of 42 patients out of 210 is not described: there is no statement of whether the split was random, stratified, or fixed by patient ID, nor any reporting of the split's class distribution or the number of tumor-bearing cases. Without this information, the local Dice values (0.969 kidney, 0.805 tumor) cannot be interpreted, and could reflect a favorable split rather than method performance. The authors should specify the split procedure and, ideally, report results across multiple split realizations.
[Section 2.4 (Inference and Post Processing) and Section 3] The connected-component postprocessing assumes 'there are at most two kidneys in one patient and the tumor should be attached with kidney.' The paper reports no analysis of how often this anatomical prior fails on the local test set or the challenge test set. If a genuine tumor or kidney fragment is disconnected from the main kidney, those true-positive voxels are discarded, which can lower tumor Dice. The reported tumor Dice of 0.805/0.8181 is likely sensitive to this choice, and without a quantitative breakdown the method's contribution relative to postprocessing remains unquantified.
[Section 3 and Section 2.3] No error bars or repeated-run variability are reported for any Dice score. The training procedure involves random patch sampling, random augmentation, and a learning-rate schedule based on the validation loss; a different random seed can shift the results. Reporting a single run for the local split and a single test-set submission does not allow the reader to assess the robustness of the claimed improvements. At minimum, a small number of repeated training runs with different seeds should be reported.

minor comments (5)

[Abstract, Section 2.3] The loss expression in Eq. (2) of Section 2.3 is missing the outer exponent on the Dice terms: the text says 'exponential logarithmic loss' and cites Wong et al., but the formula as written is not the standard exponential log Dice form (which uses exp(-x) or a nonlinear transform). Clarify the exact form or correct the equation.
[Section 2.2, Fig. 1] The paper states the actual architecture is 3D but shows a 2D schematic. This is acceptable, but the figure should clearly label the multi-scale supervision points and the corresponding resolution labels; also mention in the text the number of decoder outputs that are supervised (e.g., at which scales) and how the low-resolution labels are generated.
[Section 3, Fig. 5] The boxplot of the 42-patient test results is not accompanied by numerical quartiles or a table of per-case values. Since the manuscript reports only mean Dice, the reader cannot assess the spread for tumor, which is usually the most variable class. Include a table with median and IQR for kidney and tumor Dice.
[Section 2.1, Section 2.3] Preprocessing and training details are incomplete: the resampling target voxel spacing is not stated, the normalization is described as using the 'global foreground mean and standard deviation' but it is not indicated how foreground is defined, and the learning-rate schedule is described without the total number of epochs or the exact patience trigger. Add these details to make the method reproducible.
[Multiple places] The paper contains several typos and awkward phrasings, e.g., 'fromat' in Section 2.3, 'cant' in the abstract, 'kdney' in Section 4, and 'variance' spelled as 'varience' in Section 3. A careful proofreading pass is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the reported Dice scores are externally evaluated on KiTS19; the missing ablation weakens causal attribution but does not reduce the derivation to its inputs.

full rationale

The paper's central quantitative claim is an external benchmark result: the KiTS19 challenge score of Composite Dice 0.8961 (kidney 0.9741, tumor 0.8181) is computed by the challenge organizers on withheld test scans, and the local score of 0.969/0.805 is evaluated on a separately held-out 42-patient split. Neither score is equivalent to the training objective by construction. The training loss (Eqs. 1-3) is a Dice-plus-cross-entropy objective, and using a Dice-based loss for a Dice-evaluated segmentation task is a standard and non-circular practice: the model must still generalize to unseen patients, and the challenge metric is externally computed. The multi-scale supervision claim is presented as an architectural enhancement, but the paper does not define multi-scale supervision in terms of the final Dice, and no parameter is fitted to the test set and then renamed as a prediction. There is no load-bearing self-citation: the cited works [5], [6], [13], and [14] are external prior work, and no uniqueness theorem or ansatz is imported from the authors' own previous papers. The paper's own limitation remarks about small tumors and the absence of an ablation isolating multi-scale supervision are legitimate evidence-quality concerns, but they are not circularity. The reported composite score is internally consistent with the component Dice scores, further showing no arithmetic or definitional forcing. Therefore the derivation chain is self-contained against an external benchmark, and the appropriate circularity score is 0.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central result relies on hand-set loss weights, training hyperparameters, an anatomical prior in postprocessing, and untested design assumptions such as deep supervision helping and 5 layers being sufficient. No new physical entities are introduced; the architecture is a reconfiguration of known components, so the empirical Dice scores carry the paper.

free parameters (7)

DICE class weights = 0.4 kidney, 0.6 tumor
Equation (2) weights the soft Dice terms; the text says weights were attributed manually, and no ablation tests alternatives.
CE class weights = 0.28 background, 0.28 kidney, 0.44 tumor
Equation (3); set manually to address class imbalance, with no sensitivity analysis.
Exponential loss exponent = 0.3
Equation (2); exponent in the exponential logarithmic loss, chosen by hand.
Base feature number = 30
Section 2.2; chosen to reduce model volume.
Patch and batch size = 192x192x48, batch 8
Section 2.3; chosen because of GPU memory limits, affects training dynamics and results.
Learning rate schedule = 3e-4, decay factor 0.2, patience 30
Section 2.3; standard choices but still free hyperparameters that influence the final model.
Internal test split size = 42 of 210 patients
Section 3; arbitrary split without stratification or seed details, used for the local Dice evaluation.

assumptions (6)

domain assumption Clipping CT intensities to the 0.5 and 99.5 percentiles and normalizing by foreground mean and standard deviation is beneficial.
Section 2.1; relies on standard CT normalization practice, not validated on this dataset.
domain assumption Resampling all scans to a common voxel spacing preserves all information needed for segmentation.
Section 2.1; the paper states anisotropy would hurt 3D convolution but does not quantify any information loss.
ad hoc to paper A 5-layer U-Net is sufficient to represent the features needed for kidney and tumor classification.
Section 1; stated as an opinion from the authors' point of view, not tested against deeper variants.
ad hoc to paper Supervising every decoder resolution level improves the final full-resolution prediction.
Section 2.2; this is the core design hypothesis, asserted without an ablation.
domain assumption Anatomy prior: at most two kidneys per patient and tumor is always attached to kidney, so disconnected predictions are false.
Section 2.4; underpins postprocessing and can fail for metastatic or detached tumor tissue.
domain assumption The chosen training hyperparameters are standard and need no special justification.
Section 2.3; standard choices but still assumptions that affect the final result.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multi Scale Supervised 3D U-Net for Kidney and Tumor Segmentation." pith.science (2026). https://pith.science/paper/BM6CHOIT

@misc{pith2026190803204,
  author       = {Pith},
  title        = {Pith review of: Multi Scale Supervised 3D U-Net for Kidney and Tumor Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BM6CHOIT}},
  note         = {Machine review of arXiv:1908.03204}
}

read the original abstract

U-Net has achieved huge success in various medical image segmentation challenges. Kinds of new architectures with bells and whistles might succeed in certain dataset when employed with optimal hyper-parameter, but their generalization always can't be guaranteed. Here, we focused on the basic U-Net architecture and proposed a multi scale supervised 3D U-Net for the segmentation task in KiTS19 challenge. To enhance the performance, our work can be summarized as three folds: first, we used multi scale supervision in the decoder pathway, which could encourage the network to predict right results from the deep layers; second, with the aim to alleviate the bad effect from the sample imbalance of kidney and tumor, we adopted exponential logarithmic loss; third, a connected-component based post processing method was designed to remove the obviously wrong voxels. In the published KiTS19 training dataset (totally 210 patients), we divided 42 patients to be test dataset and finally obtained DICE scores of 0.969 and 0.805 for the kidney and tumor respectively. In the challenge, we finally achieved the 7th place among 106 teams with the Composite Dice of 0.8961, namely 0.9741 for kidney and 0.8181 for tumor.

Figures

Figures reproduced from arXiv: 1908.03204 by the authors.

**Figure 2.** The effect of our post processing method. The left image is before post processed, and the volume in dotted box are some voxels obviously wrong. The right image is after post processed, and the extra voxels have been removed. Some common human knowledge could help to enhance the performance further. For example, there are at most two kidneys in one patient and the tumor should be attached with kidney. So, we design… view at source ↗

**Figure 3.** The loss changing during our training. Red and blue lines mean validation and training loss respectively. The green line is a sliding validation loss metric to choose the best check point. The samples of our segmentation outputs are shown in Fig.4. We observed both 2D slices from different views and 3D view to analyze the performance. Obviously, the kidney in CT images was segmented pretty well. The prediction was … view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Two samples of our segmentation outputs. The rows from up to bottom are sagittal, coronal, transverse plane in 2D and the 3D view (best viewed in color). The columns from left to right are the ground truth and prediction of one common case and the worst case [PITH_FUL…

**Figure 5.** Figure 5: The boxplot of the segmentation outputs for the 42 patients in divided test dataset [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 2 canonical work pages

[1]

Kutikov, A., Uzzo, R. G. (2009). The R.E.N.A.L. Nephrometry Score: A Com- prehensive Standardized System for Quantitating Renal Tumor Size, Location and Depth. Journal of Urology. https://doi.org/10.1016/j.juro.2009.05.035

work page doi:10.1016/j.juro.2009.05.035 2009
[2]

Ficarra, V., Novara, G., Secco, S., Macchi, V., Porzionato, A., De Caro, R., Artibani, W. (2009). Preoperative Aspects and Dimensions Used for an Anatomical (PADUA) Classiﬁcation of Renal Tumours in Patients who are Candidates for Nephron- Sparing Surgery. European Urology. https://doi.org/10.1016/j.eururo.2009.07.040

work page doi:10.1016/j.eururo.2009.07.040 2009
[3]

Taha, A., Lo, P., Li, J., Zhao, T. (2018). Kid-Net: Convolution Networks for Kid- ney Vessels Segmentation from CT-Volumes. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artiﬁcial Intelligence and Lecture Notes in Bioinformatics). https://doi.org/10.1007/978-3-030-00937-3 53

doi:10.1007/978-3-030-00937-3 2018
[4]

Heller, N., Sathianathen, N., Kalapara, A., Walczak, E., Moore, K., Kaluzniak, H., Weight, C. (2019). The KiTS19 Challenge Data: 300 Kidney Tumor Cases with Clini-cal Context, CT Semantic Segmentations, and Surgical Outcomes. 113. Retrieved from http://arxiv.org/abs/1904.00445

arXiv 2019
[5]

Ronneberger, O., Fischer, P., Brox, T. (2015). U-net: Convolutional networks for bi- omedical image segmentation. Lecture Notes in Computer Science (Including Subse- ries Lecture Notes in Artiﬁcial Intelligence and Lecture Notes in Bioinformatics), 9351, 234241. https://doi.org/10.1007/978-3-319-24574-4 28

doi:10.1007/978-3-319-24574-4 2015
[6]

Isensee, F., Petersen, J., Kohl, S. A. A., Jger, P. F., Maier-Hein, K. H. (2019). nnU- Net: Breaking the Spell on Successful Medical Image Segmentation. 1, 18. Retrieved from http://arxiv.org/abs/1904.08128 8 W. Zhao and Z. Zeng

arXiv 2019
[7]

S., Brox, T., Ronneberger, O

Cicek, O., Abdulkadir, A., Lienkamp, S. S., Brox, T., Ronneberger, O. (2016). 3D UNet: Learning Dense Volumetric. Medical Image Computing and Computer- Assisted Intervention - MICCAI 2016, 424432. https://doi.org/10.1007/978-3-319- 46723-8

doi:10.1007/978-3-319- 2016
[8]

Milletari, F. (2016). V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. 2016 Fourth International Conference on 3D Vision (3DV), 565571. https://doi.org/10.1109/3DV.2016.79

doi:10.1109/3dv.2016.79 2016

Show all 14 references

[9]

W., Heng, P

Li, X., Chen, H., Qi, X., Dou, Q., Fu, C. W., Heng, P. A. (2018). H-DenseUNet: Hy- brid Densely Connected UNet for Liver and Tumor Segmentation from CT Volumes. IEEE Transactions on Medical Imaging. https://doi.org/10.1109/TMI.2018.2845918

2018
[10]

Le, Lee, M., Heinrich, M., Misawa, K., Rueck- ert, D

Oktay, O., Schlemper, J., Folgoc, L. Le, Lee, M., Heinrich, M., Misawa, K., Rueck- ert, D. (2018). Attention U-Net: Learning Where to Look for the Pancreas. (Midl). Retrieved from http://arxiv.org/abs/1804.03999

2018 arXiv
[11]

Lin, T.-Y., Dollr, P., Girshick, R., He, K., Hariharan, B., Belongie, S. (n.d.). Feature Pyramid Networks for Object Detection. Retrieved from https://arxiv.org/pdf/1612.03144.pdf

arXiv
[12]

G., Navab, N., Wachinger, C

Roy, A. G., Navab, N., Wachinger, C. (2018). Concurrent spatial and channel squeeze & excitation in fully convolutional networks. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artiﬁcial Intelligence and Lecture Notes in Bioinformatics). https://doi.or...

2018 doi
[13]

Y., Goyal, P., Girshick, R., He, K., Dollar, P

Lin, T. Y., Goyal, P., Girshick, R., He, K., Dollar, P. (2017). Focal Loss for Dense Object Detection. Proceedings of the IEEE International Conference on Computer Vision, 2017-October, 29993007. https://doi.org/10.1109/ICCV.2017.324

2017 doi
[14]

Wong, K. C. L., Moradi, M., Tang, H., Syeda-Mahmood, T. (2018). 3D seg- mentation with exponential logarithmic loss for highly unbalanced object sizes. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Ar- tiﬁcial Intelligence and Lecture Notes in Bioinfo...

2018 doi

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Kutikov, A., Uzzo, R. G. (2009). The R.E.N.A.L. Nephrometry Score: A Com- prehensive Standardized System for Quantitating Renal Tumor Size, Location and Depth. Journal of Urology. https://doi.org/10.1016/j.juro.2009.05.035

work page doi:10.1016/j.juro.2009.05.035 2009

[2] [2]

Ficarra, V., Novara, G., Secco, S., Macchi, V., Porzionato, A., De Caro, R., Artibani, W. (2009). Preoperative Aspects and Dimensions Used for an Anatomical (PADUA) Classiﬁcation of Renal Tumours in Patients who are Candidates for Nephron- Sparing Surgery. European Urology. https://doi.org/10.1016/j.eururo.2009.07.040

work page doi:10.1016/j.eururo.2009.07.040 2009

[3] [3]

Taha, A., Lo, P., Li, J., Zhao, T. (2018). Kid-Net: Convolution Networks for Kid- ney Vessels Segmentation from CT-Volumes. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artiﬁcial Intelligence and Lecture Notes in Bioinformatics). https://doi.org/10.1007/978-3-030-00937-3 53

doi:10.1007/978-3-030-00937-3 2018

[4] [4]

Heller, N., Sathianathen, N., Kalapara, A., Walczak, E., Moore, K., Kaluzniak, H., Weight, C. (2019). The KiTS19 Challenge Data: 300 Kidney Tumor Cases with Clini-cal Context, CT Semantic Segmentations, and Surgical Outcomes. 113. Retrieved from http://arxiv.org/abs/1904.00445

arXiv 2019

[5] [5]

Ronneberger, O., Fischer, P., Brox, T. (2015). U-net: Convolutional networks for bi- omedical image segmentation. Lecture Notes in Computer Science (Including Subse- ries Lecture Notes in Artiﬁcial Intelligence and Lecture Notes in Bioinformatics), 9351, 234241. https://doi.org/10.1007/978-3-319-24574-4 28

doi:10.1007/978-3-319-24574-4 2015

[6] [6]

Isensee, F., Petersen, J., Kohl, S. A. A., Jger, P. F., Maier-Hein, K. H. (2019). nnU- Net: Breaking the Spell on Successful Medical Image Segmentation. 1, 18. Retrieved from http://arxiv.org/abs/1904.08128 8 W. Zhao and Z. Zeng

arXiv 2019

[7] [7]

S., Brox, T., Ronneberger, O

Cicek, O., Abdulkadir, A., Lienkamp, S. S., Brox, T., Ronneberger, O. (2016). 3D UNet: Learning Dense Volumetric. Medical Image Computing and Computer- Assisted Intervention - MICCAI 2016, 424432. https://doi.org/10.1007/978-3-319- 46723-8

doi:10.1007/978-3-319- 2016

[8] [8]

Milletari, F. (2016). V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation. 2016 Fourth International Conference on 3D Vision (3DV), 565571. https://doi.org/10.1109/3DV.2016.79

doi:10.1109/3dv.2016.79 2016