Pith. sign in

REVIEW 5 major objections 5 minor 28 references

Bi-Directional ConvLSTM U-Net with Densley Connected Convolutions

T0 review · 5 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 U-Net variant that fuses encoder and decoder features with a bi-directional ConvLSTM and uses dense bottleneck convolutions reports higher F1, accuracy, Jaccard, and AUC than earlier U-Net-based models on retinal, skin-lesion, and…

desk verdict A plausible U-Net extension whose headline numbers are mathematically impossible, so the state-of-the-art claim does not hold up. read the letter →

arxiv 1909.00166 v1 pith:4H5UF7AN submitted 2019-08-31 eess.IV cs.CV

Reza Azad , Maryam Asadi-Aghbolaghi , Mahmood Fathy , Sergio Escalera This is my paper · ORCID
classification eess.IVcs.CV
keywords medicalimagesegmentationU-Netbi-directionalConvLSTMdenseconvolutionsbatchnormalizationretinalbloodvesselskinlesionlungnodule
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

BCDU-Net is a U-Net variant for medical image segmentation, and the paper claims it outperforms previous U-Net-based models on three public benchmarks: retinal blood vessels (DRIVE), skin lesions (ISIC 2018), and lung regions in CT scans. Its two architectural novelties are replacing the plain concatenation in U-Net's skip connections with a bi-directional ConvLSTM that fuses high-resolution encoder features with semantic decoder features, and adding densely connected convolutions to the deepest encoding block. The paper also reports that batch normalization after up-convolutions makes training converge roughly 6.6 times faster without hurting accuracy. If these claims hold, the practical upshot is that a comparatively small architectural change to a familiar backbone can raise segmentation quality and cut training time across different medical imaging modalities.

What carries the argument

The load-bearing mechanism is the bi-directional ConvLSTM (BConvLSTM) placed in every skip connection: a layer made of two ConvLSTM stacks, ConvLSTM being a recurrent unit that replaces matrix multiplications with convolutions, one scanning the input sequence forward and one backward, whose hidden states are combined by a tanh layer. The sequence has two entries, the encoded feature map and the up-convolved decoded feature map, so each feature set can modulate the other through convolutional gates instead of just being concatenated. Two supporting mechanisms carry the rest of the argument: densely connected convolution blocks in the last encoding layer, where each block receives the concatenation of all earlier blocks' outputs, and batch normalization after each up-convolution, which normalizes activations by batch mean and variance and is credited with faster convergence.

What would settle it

Re-run the three experiments with U-Net, RU-Net, R2U-Net, and BCDU-Net from one code base under identical splits, preprocessing, patch sampling, and metric definitions; if the F1/Jaccard gaps reverse or vanish, the state-of-the-art claim fails. As a direct check on the reported ISIC numbers in Table 2, recompute F1 and Jaccard from the predicted masks: for binary segmentation Jaccard cannot exceed F1, so the table's Jaccard values (0.936, 0.937) above F1 (0.847, 0.851) would have to be resolved or reproduced before that comparison can stand.

Watch

Extended reading notes

Core claim

The central claim is that the way U-Net combines encoder and decoder feature maps in the skip connection is a bottleneck, and nonlinear bidirectional fusion removes it. In BCDU-Net, the high-resolution feature maps from a contracting level and the up-sampled, semantically richer feature maps from the adjacent decoding level are fed as a two-entry sequence into a bi-directional ConvLSTM; a tanh combination of the forward and backward hidden states produces the fused map that goes to the decoder. At the deepest encoding level, a chain of densely connected convolution blocks lets every block see the concatenation of all previous blocks' outputs, encouraging feature reuse. Evaluated on DRIVE, ISIC 2018, and the Kaggle lung dataset, the paper reports F1, accuracy, Jaccard, and AUC numbers above the previously published U-Net, RU-Net, and R2U-Net baselines, with the dense version consistently better than the version without dense connections.

Load-bearing premise

The headline comparison assumes that previously published baseline numbers were produced under the same data splits, patch sampling, preprocessing, and metric definitions as the authors' runs, and that the lung-region test phase can be done without ground-truth masks.

Editorial extensions

If this is right

  • If the reported numbers hold, BCDU-Net with dense blocks is the best of the compared U-Net variants on all three datasets, not just one.
  • On ISIC 2018, the reported Jaccard climbs from 0.592 for the best attention variant to 0.937, which would mean a large practical improvement for lesion boundary delineation.
  • On the lung CT set, the reported F1 of 0.9904 and accuracy of 0.9972 would make the dense version the strongest of the compared models for this task.
  • Batch normalization after up-convolutions would cut ISIC training epochs from about 200 to about 30, a direct practical saving whenever this architecture is retrained.

Reading between the lines

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

  • A direct ablation that replaces BConvLSTM with a single ConvLSTM or an attention gate would identify whether the gain comes from bidirectionality or from learnable gating; the paper does not run that comparison.
  • Since the BConvLSTM treats the two feature sets as a length-two sequence, one can read the mechanism as mutual spatial gating: each ConvLSTM state learns to encode information relevant to the other feature set; that interpretation is not spelled out in the paper.
  • The same architectural recipe could be transplanted to other small-annotation segmentation tasks, such as microscopy or ultrasound, but the paper only evaluates fundus, dermoscopy, and CT images.
  • For real-world lung screening, the training pipeline's surrounding-mask construction uses ground truth, so a deployable system would need an automatic way to estimate the same surrounding tissue at test time; the paper sketches the idea but does not specify that estimator.
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

5 major / 5 minor

Summary. The manuscript proposes BCDU-Net, an extension of U-Net for medical image segmentation. The architecture inserts bi-directional ConvLSTM (BConvLSTM) layers into the skip connections and adds densely connected convolutional blocks in the deepest encoding layer, together with batch normalization after up-convolutions. The authors evaluate the method on three public benchmarks: DRIVE retinal vessel segmentation, ISIC 2018 skin lesion segmentation, and a Kaggle lung segmentation dataset, reporting F1-score, sensitivity, specificity, accuracy, Jaccard similarity, and AUC. The central claim is that BCDU-Net achieves state-of-the-art performance on all three datasets. The paper also includes an ablation study (Table 4) and a convergence-speed comparison with and without batch normalization.

Significance. If the reported improvements were valid, the architectural contribution would be of moderate interest to the medical image segmentation community: replacing concatenation in skip connections with BConvLSTM fusion and adding dense blocks are both plausible mechanisms for improving feature reuse, and the paper provides a public implementation link. However, the significance of the paper as submitted is undermined by internally inconsistent quantitative results. The reported F1 (Dice) values are numerically smaller than the reported Jaccard (IoU) values for the same predictions, which is mathematically impossible, and the Jaccard column appears to be a copy of the accuracy column for the proposed method. Since the headline contribution is an empirical state-of-the-art claim, these inconsistencies invalidate the central conclusion. The manuscript also leaves the lung test-time protocol underspecified because Algorithm 1 uses the ground-truth mask to construct the training surrounding mask, and the test-time analogue is not described.

major comments (5)
  1. [§4.2, Table 2 and §4.3, Table 3] The reported metrics violate a mathematical identity. For binary segmentation, F1 (Dice) is always greater than or equal to Jaccard (IoU) for the same set of predictions, with equality only in degenerate cases. Table 2 reports BCDU-Net (d=1) F1=0.847 and JS=0.936, and BCDU-Net (d=3) F1=0.851 and JS=0.937; both violate the inequality. Table 3 reports the same violation, for example BCDU-Net (d=3) F1=0.9904 and JS=0.9972. This indicates that the evaluation script or the reported numbers are unreliable, and it invalidates the state-of-the-art comparison that is the paper's central claim.
  2. [§4.2, Table 2 and §4.3, Table 3] The Jaccard column is identical to the Accuracy column for all BCDU-Net rows: Table 2 shows JS=0.936 and AC=0.936 for d=1 and JS=0.937 and AC=0.937 for d=3; Table 3 shows JS=0.9967 and AC=0.9967 for d=1 and JS=0.9972 and AC=0.9972 for d=3. Since Jaccard and accuracy are different quantities, this suggests the JS column was populated with accuracy values or the evaluation code computed the wrong metric. The authors should recompute all metrics from the raw test predictions and report the corrected tables before any performance claim can be assessed.
  3. [§4.4, Table 4 versus §4.1, Table 1] The DRIVE results are inconsistent between the two tables. Table 1 reports BCDU-Net (d=3) F1=0.8224, while Table 4 reports the same configuration as F1=0.8243. Likewise the U-Net accuracy values differ between Table 4 and Table 3 for the lung dataset (AC=0.9828 in Table 4 versus AC=0.9872 in Table 3). The authors need to explain which numbers come from which experimental protocol, and the final tables must report a single consistent set of test-set results, ideally with standard deviations over multiple runs.
  4. [§4.3, Algorithm 1] Algorithm 1 takes both the input X and the ground-truth mask GT Mask as inputs and constructs a surrounding mask by subtracting the ground truth. The text says this preprocessing is applied to the training set, but the test-time protocol is not specified. If the lung-region estimate at test time relies on the same surrounding-mask construction, it would require ground-truth masks on the test set, which makes the evaluation circular. The authors must describe how the surrounding mask is obtained for test images without ground-truth annotations and justify that the training and test protocols are aligned.
  5. [§4.1–§4.3, Tables 1–3] The state-of-the-art comparison is performed by quoting numbers from prior papers rather than by re-running baseline methods under the authors' exact training protocol, preprocessing, and split. Given that the metric errors above already show the authors' evaluation pipeline is not reliable, the cross-paper baseline comparisons cannot be accepted at face value. At minimum, the authors should re-implement U-Net, RU-Net, and R2U-Net under identical conditions, or clearly state the protocol differences and provide the uncertainty of each reported number.
minor comments (5)
  1. [Title] The title contains a typo: 'Densley' should be 'Densely'.
  2. [§4, Experimental Results] The text says 'Keras with TenserFlow backend'; 'TenserFlow' should be 'TensorFlow', and 'datase' in §4.1 should be 'dataset'.
  3. [§4.4, Discussion] The sentence 'ia able to encode relevant information' contains a typo: 'ia' should be 'is'.
  4. [Figure 3 caption] The caption of Figure 3 says 'Bi-directional ConvLSTM in CUA-Net', but the network is called BCDU-Net; the caption should be corrected.
  5. [§3.2.2, Bi-Directional ConvLSTM] The notation for the ConvLSTM equations is inconsistent: Equation (1) uses Wci, Wcf, and Wco with a Hadamard product in the cell-update and output-gate terms, whereas standard ConvLSTM formulations use convolution operations throughout; the authors should clarify the exact tensor operations and dimensions.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all load-bearing claims are empirical comparisons on public benchmarks against external baselines.

full rationale

The paper proposes an architectural modification of U-Net (BConvLSTM in skip connections, densely connected convolutions in the deepest encoding block, and batch normalization after up-convolutions) and supports its claims by direct experiments on three public datasets. There is no derivation from fitted parameters that is then renamed as a prediction: the reported F1, accuracy, Jaccard, sensitivity, specificity, and AUC values are measured outputs of trained networks, not functions of a fitted model. The BConvLSTM equations and dense-block description are presented as standard components from the cited literature; the paper does not claim to derive them from first principles, nor does it define any quantity in terms of its own output. The choice d=1 versus d=3 is a hyperparameter comparison reported on the same benchmarks, which is ordinary model selection rather than a circular prediction; the paper does not fit a parameter to a subset and then predict a closely related derived quantity. Baselines such as U-Net, RU-Net, R2U-Net, and Attention U-Net are taken from independently authored prior work, and the paper's citations of Alom et al. are not self-citations. Any concern about inconsistent evaluation protocols between the paper's runs and the quoted baselines is a correctness and reproducibility issue, not circularity, because the central claim's support is an external benchmark comparison rather than a self-referential derivation. The internally inconsistent F1 and Jaccard values in Tables 2 and 3 indicate an evaluation-script or reporting problem, not a case where the result is assumed in the input. Therefore, under the required evidentiary standard, no circular step can be exhibited.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central claim is an empirical performance claim. The only manually chosen free parameter is the number of dense blocks, d=3. The axioms are the assumptions about comparability of baselines, the test-time lung procedure, and the correctness of metric computations.

free parameters (1)
  • Number of dense blocks (d) = 3
    Manually chosen; the paper compares d=1 and d=3 and adopts d=3 for the main results. No search or sensitivity analysis beyond these two values is reported.
assumptions (3)
  • domain assumption Baseline numbers from prior publications are directly comparable to the authors' runs.
    Section 4 compares BCDU-Net with U-Net, RU-Net, and R2U-Net using values taken from cited papers; no controlled reimplementation is described.
  • domain assumption Lung surrounding-tissue masks can be estimated at test time without ground-truth masks.
    Algorithm 1 in Section 4.3 takes the ground-truth mask as input to build the surrounding mask, but the test-time procedure for estimating the surrounding region is not specified.
  • domain assumption The reported F1 and Jaccard metrics are computed on the same binary predictions.
    The tables in Section 4 list F1 and JS values; for the same binary masks, Dice is always at least IoU, so the reported values imply a computation error unless different thresholds or predictions were used, which is not stated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bi-Directional ConvLSTM U-Net with Densley Connected Convolutions." pith.science (2026). https://pith.science/paper/4H5UF7AN

@misc{pith2026190900166,
  author       = {Pith},
  title        = {Pith review of: Bi-Directional ConvLSTM U-Net with Densley Connected Convolutions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4H5UF7AN}},
  note         = {Machine review of arXiv:1909.00166}
}
read the original abstract

In recent years, deep learning-based networks have achieved state-of-the-art performance in medical image segmentation. Among the existing networks, U-Net has been successfully applied on medical image segmentation. In this paper, we propose an extension of U-Net, Bi-directional ConvLSTM U-Net with Densely connected convolutions (BCDU-Net), for medical image segmentation, in which we take full advantages of U-Net, bi-directional ConvLSTM (BConvLSTM) and the mechanism of dense convolutions. Instead of a simple concatenation in the skip connection of U-Net, we employ BConvLSTM to combine the feature maps extracted from the corresponding encoding path and the previous decoding up-convolutional layer in a non-linear way. To strengthen feature propagation and encourage feature reuse, we use densely connected convolutions in the last convolutional layer of the encoding path. Finally, we can accelerate the convergence speed of the proposed network by employing batch normalization (BN). The proposed model is evaluated on three datasets of: retinal blood vessel segmentation, skin lesion segmentation, and lung nodule segmentation, achieving state-of-the-art performance.

Figures

Figures reproduced from arXiv: 1909.00166 by the authors.

Figure 1
Figure 1. The network utilizes the strengths of both BCon [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 1
Figure 1. BCDU-Net with bi-directional ConvLSTM in the skip connections and densely connected convolution. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Bi-directional ConvLSTM in CUA-Net. where ∗ and ◦ denote the convolution and Hadamard func￾tions, respectively. Xt is the input tensor (in our case Xe and Xbup d ), Ht is the hidden sate tensor, Ct is the memory cell tensor, and, Wx∗ and Wh∗ are 2D Convolution kernels corresponding to the input and hidden state, respectively, and bi , bf , bo, and bc are the bias terms. In this network, we employ BConvLSTM [23] to e… view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Segmentation result of BCDU-Net on DRIVE. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Training and validation accuracy of BCDU-Net for three datasets. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: ROC diagrams of the proposed BCDU-Net for three dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Segmentation result of BCDU-Net on ISIC. [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 9
Figure 9. Figure 9: Segmentation result of BCDU-Net on Lung dataset. [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]
Figure 10
Figure 10. Figure 10: Visual effect of BConvLSTM in BCDU-Net. We included BN after each up-convolutional layer to speed up the network learning process. To evaluate the ef￾fect of this function, we train the network with and without BN [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]
Figure 11
Figure 11. Figure 11: Training and validation accuracy of BCDU-Net (a) [PITH_FULL_IMAGE:figures/full_fig_p008_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 22 canonical work pages

  1. [1]

    https://www.kaggle.com/kmader/ finding-lungs-in-ct-data

    work page

  2. [2]

    M. Z. Alom, M. Hasan, C. Yakopcic, T. M. Taha, and V . K. Asari. Recurrent residual convolutional neural net- work based on u-net (r2u-net) for medical image segmen- tation. arXiv preprint arXiv:1802.06955, 2018

    arXiv 2018

  3. [3]

    Azzopardi, N

    G. Azzopardi, N. Strisciuglio, M. Vento, and N. Petkov. Trainable cosfire filters for vessel delineation with applica- tion to retinal images. Medical image analysis, 19(1):46–57, 2015

    work page 2015

  4. [4]

    W. Bai, H. Suzuki, C. Qin, G. Tarroni, O. Oktay, P. M. Matthews, and D. Rueckert. Recurrent neural networks for aortic image sequence segmentation with sparse annota- tions. In International Conference on Medical Image Com- puting and Computer-Assisted Intervention , pages 586–594. Springer, 2018

    work page 2018

  5. [5]

    H. Chen, Q. Dou, L. Yu, J. Qin, and P.-A. Heng. V oxres- net: Deep voxelwise residual networks for brain segmenta- tion from 3d mr images. NeuroImage, 170:446–455, 2018

    work page 2018

  6. [6]

    L.-C. Chen, G. Papandreou, I. Kokkinos, K. Murphy, and A. L. Yuille. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully con- nected crfs. IEEE transactions on pattern analysis and ma- chine intelligence, 40(4):834–848, 2017

    work page 2017

  7. [7]

    C ¸ ic ¸ek, A

    ¨O. C ¸ ic ¸ek, A. Abdulkadir, S. S. Lienkamp, T. Brox, and O. Ronneberger. 3d u-net: learning dense volumetric seg- mentation from sparse annotation. In International confer- ence on medical image computing and computer-assisted in- tervention, pages 424–432. Springer, 2016

    work page 2016

  8. [8]

    N. C. Codella, D. Gutman, M. E. Celebi, B. Helba, M. A. Marchetti, S. W. Dusza, A. Kalloo, K. Liopyris, N. Mishra, H. Kittler, et al. Skin lesion analysis toward melanoma de- tection: A challenge at the 2017 international symposium on biomedical imaging (isbi), hosted by the international skin imaging collaboration (isic). In ISBI 2018, pages 168–172. IEEE, 2018

    work page 2017

Show all 28 references
  1. [9]

    Z. Cui, R. Ke, and Y . Wang. Deep bidirectional and unidirec- tional lstm recurrent neural network for network-wide traffic speed prediction. arXiv preprint arXiv:1801.02143, 2018

    2018 arXiv

  2. [10]

    Z. Cui, J. Yang, and Y . Qiao. Brain mri segmentation with patch-based cnn approach. In 2016 35th Chinese Control Conference (CCC), pages 7026–7031. IEEE, 2016

    2016

  3. [11]

    Drozdzal, E

    M. Drozdzal, E. V orontsov, G. Chartrand, S. Kadoury, and C. Pal. The importance of skip connections in biomedical image segmentation. In Deep Learning and Data Labeling for Medical Applications, pages 179–187. Springer, 2016

    2016

  4. [12]

    Huang, Z

    G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Wein- berger. Densely connected convolutional networks. In Pro- ceedings of the IEEE conference on computer vision and pat- tern recognition, pages 4700–4708, 2017

    2017

  5. [13]

    Ioffe and C

    S. Ioffe and C. Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift. pages 448–456, 2015

    2015

  6. [14]

    Kleesiek, G

    J. Kleesiek, G. Urban, A. Hubert, D. Schwarz, K. Maier- Hein, M. Bendszus, and A. Biller. Deep mri brain extrac- tion: a 3d convolutional neural network for skull stripping. NeuroImage, 129:460–469, 2016

    2016

  7. [15]

    Q. Li, B. Feng, L. Xie, P. Liang, H. Zhang, and T. Wang. A cross-modality learning approach for vessel segmentation in retinal images. IEEE transactions on medical imaging , 35(1):109–118, 2015

    2015

  8. [16]

    Liskowski and K

    P. Liskowski and K. Krawiec. Segmenting retinal blood ves- sels with deep neural networks. IEEE transactions on medi- cal imaging, 35(11):2369–2380, 2016

    2016

  9. [17]

    J. Long, E. Shelhamer, and T. Darrell. Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 3431–3440, 2015

    2015

  10. [18]

    Milletari, N

    F. Milletari, N. Navab, and S.-A. Ahmadi. V-net: Fully convolutional neural networks for volumetric medical image segmentation. In 2016 F ourth International Conference on 3D Vision (3DV), pages 565–571. IEEE, 2016

    2016

  11. [19]

    Oktay, J

    O. Oktay, J. Schlemper, L. L. Folgoc, M. Lee, M. Hein- rich, K. Misawa, K. Mori, S. McDonagh, N. Y . Hammerla, B. Kainz, et al. Attention u-net: Learning where to look for the pancreas. arXiv preprint arXiv:1804.03999, 2018

    2018 arXiv

  12. [20]

    P. O. Pinheiro and R. Collobert. Recurrent convolutional neural networks for scene labeling. Technical report, 2014

    2014

  13. [21]

    Ronneberger, P

    O. Ronneberger, P. Fischer, and T. Brox. U-net: Convo- lutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention , pages 234–241. Springer, 2015

    2015

  14. [22]

    H. R. Roth, L. Lu, A. Farag, H.-C. Shin, J. Liu, E. B. Turk- bey, and R. M. Summers. Deeporgan: Multi-level deep convolutional networks for automated pancreas segmenta- tion. In International conference on medical image com- puting and computer-assisted intervention , pages 55...

    2015

  15. [23]

    H. Song, W. Wang, S. Zhao, J. Shen, and K.-M. Lam. Pyra- mid dilated deeper convlstm for video salient object detec- tion. In Proceedings of the European Conference on Com- puter Vision (ECCV), pages 715–731, 2018

    2018

  16. [24]

    Staal, M

    J. Staal, M. D. Abr `amoff, M. Niemeijer, M. A. Viergever, and B. Van Ginneken. Ridge-based vessel segmentation in color images of the retina. IEEE transactions on medical imaging, 23(4):501–509, 2004

    2004

  17. [25]

    Visin, M

    F. Visin, M. Ciccone, A. Romero, K. Kastner, K. Cho, Y . Bengio, M. Matteucci, and A. Courville. Reseg: A recur- rent neural network-based model for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, pages 41–48, 2016

    2016

  18. [26]

    Xingjian, Z

    S. Xingjian, Z. Chen, H. Wang, D.-Y . Yeung, W.-K. Wong, and W.-c. Woo. Convolutional lstm network: A machine learning approach for precipitation nowcasting. In Advances in neural information processing systems , pages 802–810, 2015

    2015

  19. [27]

    X. Zhou, T. Ito, R. Takayama, S. Wang, T. Hara, and H. Fu- jita. Three-dimensional ct image segmentation by combining 2d fully convolutional network with 3d majority voting. In Deep Learning and Data Labeling for Medical Applications , pages 111–120. Springer, 2016

    2016

  20. [28]

    Y . Zhou, L. Xie, W. Shen, Y . Wang, E. K. Fishman, and A. L. Yuille. A fixed-point model for pancreas segmentation in abdominal ct scans. In International Conference on Medi- cal Image Computing and Computer-Assisted Intervention , pages 693–701. Springer, 2017

    2017

Pith tools

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