Pith. sign in

REVIEW 4 major objections 7 minor 2 cited by

KPIs 2024 Challenge: Advancing Glomerular Segmentation from Patch- to Slide-Level

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

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

Pith's one-line read Kidney-spotting benchmarks cross 94 percent Dice on whole slide images.

desk verdict A genuinely useful CKD glomerular benchmark with a transparent evaluation, but the dataset size claim needs correcting and the WSI leaderboard rests on 12 test slides. read the letter →

arxiv 2502.07288 v1 pith:TVK3JWQM submitted 2025-02-11 cs.CV cs.AI

Ruining Deng , Tianyuan Yao , Yucheng Tang , Junlin Guo , Siqi Lu , Juming Xiong , Lining Yu , Quan Huu Cap
show 39 more authors
Pengzhou Cai Libin Lan Ze Zhao Adrian Galdran Amit Kumar Gunjan Deotale Dev Kumar Das Inyoung Paik Joonho Lee Geongyu Lee Yujia Chen Wangkai Li Zhaoyang Li Xuege Hou Zeyuan Wu Shengjin Wang Maximilian Fischer Lars Kramer Anghong Du Le Zhang Maria Sanchez Sanchez Helena Sanchez Ulloa David Ribalta Heredia Carlos Perez de Arenaza Garcia Shuoyu Xu Bingdou He Xinping Cheng Tao Wang Noemie Moreau Katarzyna Bozek Shubham Innani Ujjwal Baid Kaura Solomon Kefas Bennett A. Landman Yu Wang Shilin Zhao Mengmeng Yin Haichun Yang Yuankai Huo
This is my paper · ORCID
classification cs.CVcs.AI
keywords glomerularsegmentationchronickidneydiseasewholeslideimagingdeeplearningbenchmarkPAS-stainedhistologyDicesimilaritycoefficientpathologychallengerodentmodels
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 reports a community benchmark for segmenting glomeruli in kidney pathology, built from preclinical mouse models of chronic kidney disease rather than only healthy tissue. It introduces a dataset with over 10,000 annotated glomeruli from PAS-stained whole slide images spanning normal, 5/6-nephrectomy, diabetic nephropathy, and NEP25 conditions, and it runs two tasks: patch-level segmentation and whole-slide segmentation and detection. The paper positions this as the first challenge dedicated exclusively to segmenting functional units in kidney pathology across diverse CKD models. The strongest submitted methods reach Dice scores above 94% at both scales, with the best patch-level entry at 94.51 ± 6.71 and the best slide-level segmentation at 94.64 ± 0.89.

What carries the argument

The load-bearing object is the challenge itself: a curated set of PAS-stained whole slide images with pathologist-drawn glomerulus masks, split into training, validation, and private test sets, plus a fixed evaluation protocol using the Dice Similarity Coefficient for segmentation and the F1 score for detection. The mechanism that carries the argument is the two-task design: task 1 isolates local segmentation ability on 2048×2048 patches, and task 2 forces methods to stitch patch predictions back into gigapixel slide-level masks under boundary effects. The winning approaches tend to combine multiple pretrained transformer or hybrid encoders, overlap the inference windows, and sum or crop-and-paste predictions in intersecting regions before thresholding. This consistency is what lets the paper draw conclusions about which design choices matter when moving from patch scale to slide scale.

What would settle it

Re-annotate the 12 test whole-slide images by a second, independent set of pathologists and recompute the task-2 leaderboard; if several teams’ ranks move by more than one position, or if inter-observer Dice is below about 0.90, the reliability premise is violated. A simpler check is to report per-CKD-model Dice separately and see whether the top method’s 94.64 mean hides a much lower score on one model such as NEP25 or 5/6Nx.

Watch

Extended reading notes

Core claim

The central claim is that a challenge organized around this dataset can serve as a reference benchmark for glomerular segmentation across diverse chronic kidney disease models. The paper argues that previous public kidney-pathology resources mostly contain healthy samples, and that this dataset fills the gap by adding whole kidney sections from diseased rodents. On the benchmark, top methods are close together in patch-level Dice yet spread more widely at slide level, which the paper interprets as evidence that ensembling and post-processing determine whole-slide robustness. The strongest segmentation result is a mean Dice of 94.64 ± 0.89 at whole-slide level, while the strongest detection result is an F1 score of 91.33 ± 3.84.

Load-bearing premise

The entire leaderboard rests on the assumption that the 12 private test whole-slide images and their pathologist annotations fairly represent the four chronic kidney disease models; no inter-observer agreement is reported, so if those images or annotations are not representative, the rankings and the claim of robust performance across diverse models could change.

Editorial extensions

If this is right

  • If the leaderboard reflects true performance, then deep segmentation models trained on these four mouse models can generalize across CKD conditions without per-model retraining.
  • Patch-level segmentation is nearly saturated among top entries, so gains at slide level now come from ensembling, stitching, and detection-aware post-processing rather than from new network backbones alone.
  • The private test set and released code make future methods directly comparable against these results on the same disease-model distributions.
  • The challenge leaderboard gives a concrete reference for achievable slide-level glomerular detection performance, around 90% F1, and thereby a baseline for clinical-quantification tools.

Reading between the lines

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

  • If the same patch models were evaluated on human biopsies or on other stains such as H&E, the 94% Dice would likely drop; the dataset covers only PAS-stained mouse kidneys, so “diverse CKD models” currently means diverse mouse models, not diverse species or histology protocols.
  • A natural next test is to measure inter-observer agreement among the three pathologists on a subset of the whole-slide masks; the benchmark does not report it, and if human agreement is itself below the algorithm spread, leaderboard differences of a few Dice points may be within annotation noise.
  • A testable extension is to train a detection-only model on the released training set and compare its F1 against the top detection entry, which would isolate how much of the detection gap comes from segmentation architecture versus post-processing.
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 reports the KPIs 2024 MICCAI challenge for glomerular segmentation in Periodic Acid Schiff (PAS)-stained mouse kidney whole slide images. It presents two tasks: patch-level glomeruli segmentation and WSI-level segmentation/detection, evaluated with Dice Similarity Coefficient and F1 score. The dataset includes four mouse models (normal, 5/6Nx, diabetic nephropathy, and NEP25), with all glomeruli annotated by three pathologists. The paper gives the top-10 leaderboards for each task, describes the top-3 methods in each task, and provides a supplementary challengeR-based analysis of ranking stability. The central claims are that this is the first MICCAI challenge exclusively dedicated to segmenting functional units in kidney pathology across diverse CKD models, that the dataset contains over 10,000 annotated glomeruli from 60+ WSIs, and that top methods reach Dice scores of 94.51±6.71 on patch-level segmentation (Task 1) and 94.64±0.89 on WSI-level segmentation (Task 2a).

Significance. If the dataset-scale claims and leaderboard are taken at face value, the challenge fills a real gap: public kidney-pathology segmentation benchmarks are scarce, and the inclusion of four rodent CKD models with whole-slide annotations is valuable for method comparison and for stimulating work on slide-level generalization. The paper's strengths are its transparent evaluation protocol (private test set, Docker-based submissions on fixed hardware, public leaderboard, participant code links) and the supplementary statistical analysis using challengeR, including bootstrap ranking stability, significance maps, and ranking-method robustness. Because the benchmark compares external participant algorithms rather than fitting parameters to a desired conclusion, the evaluation is not circular. The main comparative claims should be read as evidence about this specific challenge rather than as a general state-of-the-art statement; after correcting the dataset-inventory discrepancy and adding annotation-reliability information, this would be a useful community resource.

major comments (4)
  1. [Abstract, Section 1, Table 1] The Abstract and Section 1 state that the dataset contains 'over 10,000 annotated glomeruli from 60+ PAS-stained whole slide images,' but Table 1 lists 5+5+5+10=25 training WSIs, 8 validation WSIs, and 12 test WSIs, totaling 45 WSIs. The Task 2 benchmarking report in the supplementary materials also analyzes 12 test cases, consistent with Table 1. The 60+ figure is therefore contradicted by the paper's own tabulation. Either Table 1 omits WSIs or the abstract and introduction overstate the dataset by one-third. Because the dataset scale is a central contribution claim, the discrepancy must be resolved before publication, and the '10,000 glomeruli' count should be reconciled with the WSI and patch counts.
  2. [Section 3.2, Section 5] Section 3.2 states that all glomeruli were annotated by three experienced pathologists using QuPath, but no inter-observer agreement is reported in the main text or the supplementary materials. This matters because Task 2's test set consists of only 12 WSIs and the top Task 1 teams differ by only 0.2-0.3 Dice points; annotation variability could be comparable to or larger than the observed method differences. Please report an agreement measure (e.g., pairwise Dice or Fleiss' kappa) on a subset of slides, describe how consensus masks were formed, and if an agreement study is unavailable, state this limitation explicitly when making comparative leaderboard claims.
  3. [Section 3.4, Table 4] The Task 2a overall ranking is described as combining DSC and F1 'equally,' but the text does not specify whether ranks or raw scores are averaged, how ties are resolved, or why Zhijian Life and Deep Bio are both assigned rank 2 while Table 4 lists them in a particular order. Without this aggregation formula, the leaderboard is not reproducible from Tables 4 and 5 alone. Please state the exact formula (e.g., mean of ranks versus rank of means) and provide the per-team rank inputs used to compute the overall rank.
  4. [Supplementary §3.3, Section 5.1] The supplementary significance maps show that many pairwise differences between adjacent leaderboard positions are not statistically significant after Holm correction, yet Sections 5.1 and 6 describe teams as 'leads,' 'follows closely,' and 'more stable' without acknowledging this. Please report which pairwise differences are actually significant and temper the comparative wording accordingly, especially for the 0.2-0.3-point gaps among the top Task 1 teams.
minor comments (7)
  1. [Section 3] The section heading '3. Challege description' contains a typo and should read 'Challenge description.'
  2. [Section 2.2] The phrase 'computational kidnet pathology' should be 'computational kidney pathology.'
  3. [Table 2] In the footnote for the Aira Matrix augmentation set, 'sptail scaling' appears to be a typo for 'spatial scaling.'
  4. [Section 6] The sentence 'A recurring trend among high-performing methods was the athe use of transformers or hybrid architectures' contains a stray 'the athe' and should be corrected.
  5. [Section 5.2.2] The text 'CV AILABand Zhijian Life' is missing a space between 'CV AILAB' and 'and.'
  6. [Figure 3 caption] The caption describes the figure as 'Patch-level quantitative results,' but the figure shows qualitative segmentation overlays; the caption should be revised or the figure should include the quantitative summary it claims to show.
  7. [Table 1] The NEP25 validation row lists an optical magnification of 40x with 0.25 µm/px resolution, while the NEP25 training row lists 20x optical magnification with 0.24 µm/px; please verify these values, as the same resolution would be expected to correspond to the same optical magnification.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the challenge leaderboard derives from external participant submissions evaluated on withheld test data; dataset inconsistencies are reporting concerns, not circular reasoning.

full rationale

This paper is a MICCAI challenge report rather than a derivation. Its central outputs—the Task 1 and Task 2 leaderboards (Tables 3–5)—are computed from participant Docker submissions evaluated on a withheld test set, so the reported Dice and F1 scores are external to the paper's own fitted values or definitions. The metrics (DSC, F1) are standard and are defined in Section 3.4 independently of any participant's model, and no parameter is fitted to a subset of data and then renamed as a prediction. The many self-citations in Section 2 (e.g., Glo-In-One, GloFinder, WCF) are contextual literature review and are not load-bearing for the challenge's ranking claims; the rankings are produced by the external challengeR toolkit described in the supplementary benchmarking report. The only mild sense in which the organizers 'define both the data and the evaluation metric' is inherent to any challenge and does not make the participants' measured performance equivalent to the input data. The discrepancy between the abstract's '60+ WSIs' and Table 1's total of 45 WSIs, and the absence of reported inter-observer agreement, are data-quality and reporting issues that affect correctness and interpretability, but they do not constitute a circular step under the definitions used here. No derivation chain reduces to its own inputs, so the appropriate score is 0.

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

The paper's results are empirical benchmark outcomes, not derivations, so no fitted parameters or invented entities are introduced. The main assumptions are domain-level, concerning annotation validity, metric sufficiency, and model diversity.

assumptions (3)
  • domain assumption PAS-stained rodent kidney sections annotated by three pathologists are a valid gold standard for glomerular segmentation.
    Invoked in Section 3.2.1 and Section 3.4.1; no inter-observer variability is reported.
  • domain assumption Dice similarity coefficient and F1 score are sufficient metrics for both segmentation tasks.
    Defined in Section 3.4.1; no boundary-based or object-level metrics are used, which could bias results toward larger glomeruli.
  • domain assumption The four mouse models (normal, 5/6Nx, DN, NEP25) represent diverse CKD tissue conditions for benchmarking.
    Stated in Sections 1 and 3.2.1; the authors acknowledge in Section 6 that more interspecies data and stains are still needed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of KPIs 2024 Challenge: Advancing Glomerular Segmentation from Patch- to Slide-Level." pith.science (2026). https://pith.science/paper/TVK3JWQM

@misc{pith2026250207288,
  author       = {Pith},
  title        = {Pith review of: KPIs 2024 Challenge: Advancing Glomerular Segmentation from Patch- to Slide-Level},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TVK3JWQM}},
  note         = {Machine review of arXiv:2502.07288}
}
read the original abstract

Chronic kidney disease (CKD) is a major global health issue, affecting over 10% of the population and causing significant mortality. While kidney biopsy remains the gold standard for CKD diagnosis and treatment, the lack of comprehensive benchmarks for kidney pathology segmentation hinders progress in the field. To address this, we organized the Kidney Pathology Image Segmentation (KPIs) Challenge, introducing a dataset that incorporates preclinical rodent models of CKD with over 10,000 annotated glomeruli from 60+ Periodic Acid Schiff (PAS)-stained whole slide images. The challenge includes two tasks, patch-level segmentation and whole slide image segmentation and detection, evaluated using the Dice Similarity Coefficient (DSC) and F1-score. By encouraging innovative segmentation methods that adapt to diverse CKD models and tissue conditions, the KPIs Challenge aims to advance kidney pathology analysis, establish new benchmarks, and enable precise, large-scale quantification for disease research and diagnosis.

Figures

Figures reproduced from arXiv: 2502.07288 by the authors.

Figure 1
Figure 1. Overview of the challenge task setup. The KPIs challenge involves two tasks: (1) Patch-level Diseased Glomeruli Segmentation, focusing on the precise delineation of glomeruli in high-resolution image patches, and (2) Whole Slide Image (WSI)-level Diseased Glomeruli Segmentation, requiring robust segmentation across entire tissue sections. Example images illustrate the input data, ground truth masks, and overlay visu… view at source ↗
Figure 2
Figure 2. Box plots of quantitative results on different tasks from participants. This figure shows the box plots of the quantitative performance of participants on two tasks. Results from the top 10 teams are reported. Dice similarity coefficient scores are provided for Task 1, while both Dice similarity coefficient scores and F1 scores are presented for Task 2 [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Patch-level quantitative results - The segmentation results from different participants’ methods on glomeruli under various experimental conditions (normal, NEP25, DN, and 5/6Nx). glomeruli samples. Top-performing teams (left columns) han￾dle boundary segmentation more effectively than lower-ranked teams (right columns), as seen in the 5/6Nx and NEP25 rows. A challenge in this task is the large size of the given ima… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: WSI-level quantitative results - This figure shows the segmentation results from different participants’ methods on glomeruli under various experimental conditions (normal, NEP25, DN, and 5/6Nx). The yellow bounding box highlights the zoomed-in region on the WSI, with …

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Controllable Diffusion-Based Lesion Inpainting for Scalable Histopathology Data Augmentation

    cs.CV 2026-01 conditional novelty 6.0 of 10

    PathoGen generates realistic lesion inpaintings in four histopathology datasets and improves downstream segmentation Dice scores by up to ~0.18–0.24 when used as augmentation.

  2. CASC-AI: Consensus-aware Self-corrective Learning for Noise Cell Segmentation

    cs.CV 2025-02 conditional novelty 5.0 of 10

    A consensus-aware self-corrective loss weighting method improves noisy-label cell segmentation on one real and two simulated glomerular datasets, but the false-positive correction claim is not consistently supported.

Reference graph

Works this paper leans on

62 extracted references · 52 canonical work pages · cited by 2 Pith papers

  1. [1]

    , author Sethi, S

    author Agarwal, S. , author Sethi, S. , author Dinda, A. , year 2013 . title Basics of kidney biopsy: A nephrologist's perspective . journal Indian journal of nephrology volume 23 , pages 243--252

    work page 2013

  2. [2]

    , author Zhong, J

    author Al Tuhaifi, T. , author Zhong, J. , author Yang, H.C. , author Fogo, A.B. , year 2024 . title Effects of dipeptidyl peptidase-4 inhibitor and angiotensin-converting enzyme inhibitor on experimental diabetic kidney disease . journal Laboratory Investigation volume 104 , pages 100305

    work page 2024

  3. [3]

    , author Cascarano, G.D

    author Altini, N. , author Cascarano, G.D. , author Brunetti, A. , author Marino, F. , author Rocchetti, M.T. , author Matino, S. , author Venere, U. , author Rossini, M. , author Pesce, F. , author Gesualdo, L. , author Bevilacqua, V. , year 2020 . title Semantic segmentation framework for glomeruli detection and classification in kidney histological sec...

    work page doi:10.3390/electronics9030503 2020

  4. [4]

    , author Loughrey, M.B

    author Bankhead, P. , author Loughrey, M.B. , author Fern \'a ndez, J.A. , author Dombrowski, Y. , author McArt, D.G. , author Dunne, P.D. , author McQuaid, S. , author Gray, R.T. , author Murray, L.J. , author Coleman, H.G. , et al., year 2017 . title Qupath: Open source software for digital pathology image analysis . journal Scientific reports volume 7 ...

    work page 2017

  5. [5]

    , author Mayerich, D

    author Becker, J.U. , author Mayerich, D. , author Padmanabhan, M. , author Barratt, J. , author Ernst, A. , author Boor, P. , author Cicalese, P.A. , author Mohan, C. , author Nguyen, H.V. , author Roysam, B. , year 2020 . title Artificial intelligence and machine learning in nephropathology . journal Kidney International volume 98 , pages 65--75 . https...

    work page doi:10.1016/j.kint.2020.02.027 2020

  6. [6]

    , author Hermsen, M

    author Bel, T.d. , author Hermsen, M. , author Litjens, G. , author Laak, J. , year 2018 . title Structure instance segmentation in renal tissue: a case study on tubular immune cell detection , in: booktitle Computational Pathology and Ophthalmic Medical Image Analysis . publisher Springer , pp. pages 112--119

    work page 2018

  7. [7]

    , author Danielsen, H

    author Bengtsson, E. , author Danielsen, H. , author Treanor, D. , author Gurcan, M.N. , author MacAulay, C. , author Moln \'a r, B. , year 2017 . title Computer-aided diagnostics in digital pathology

    work page 2017

  8. [8]

    , author Bougleux, S

    author Brixtel, R. , author Bougleux, S. , author Lézoray, O. , author Caillot, Y. , author Lemoine, B. , author Fontaine, M. , author Nebati, D. , author Renouf, A. , year 2022 . title Whole slide image quality in digital pathology: Review and perspectives . journal IEEE Access volume 10 , pages 131005--131035 . :10.1109/ACCESS.2022.3227437

    work page arXiv 2022

Show all 62 references
  1. [9]

    , author Fernandez-Carrobles, M.M

    author Bueno, G. , author Fernandez-Carrobles, M.M. , author Gonzalez-Lopez, L. , author Deniz, O. , year 2020 . title Glomerulosclerosis identification in whole slide images using semantic segmentation . journal Computer Methods and Programs in Biomedicine volume 184 , pages ...

    2020

  2. [10]

    , author Innocenti, S

    author Cirillo, L. , author Innocenti, S. , author Becherucci, F. , year 2024 . title Global epidemiology of kidney cancer . journal Nephrology Dialysis Transplantation volume 39 , pages 920--928

    2024

  3. [11]

    , author Deng, R

    author Cui, C. , author Deng, R. , author Guo, J. , author Liu, Q. , author Yao, T. , author Yang, H. , author Huo, Y. , . title Enhancing physician flexibility: Prompt-guided multi-class pathological segmentation for diverse outcomes , in: booktitle IEEE-EMBS International Co...

  4. [12]

    , author Cui, C

    author Deng, R. , author Cui, C. , author Liu, Q. , author Yao, T. , author Remedios, L.W. , author Bao, S. , author Landman, B.A. , author Wheless, L.E. , author Coburn, L.A. , author Wilson, K.T. , et al., year 2023 a. title Segment anything model (sam) for digital pathology...

    2023 arXiv

  5. [13]

    , author Liu, Q

    author Deng, R. , author Liu, Q. , author Cui, C. , author Yao, T. , author Long, J. , author Asad, Z. , author Womick, R.M. , author Zhu, Z. , author Fogo, A.B. , author Zhao, S. , et al., year 2023 b. title Omni-seg: A scale-aware dynamic network for renal pathological image...

    2023

  6. [14]

    , author Liu, Q

    author Deng, R. , author Liu, Q. , author Cui, C. , author Yao, T. , author Xiong, J. , author Bao, S. , author Li, H. , author Yin, M. , author Wang, Y. , author Zhao, S. , et al., year 2024 a. title Hats: Hierarchical adaptive taxonomy segmentation for panoramic pathology im...

    2024

  7. [15]

    , author Liu, Q

    author Deng, R. , author Liu, Q. , author Cui, C. , author Yao, T. , author Yue, J. , author Xiong, J. , author Yu, L. , author Wu, Y. , author Yin, M. , author Wang, Y. , et al., year 2024 b. title Prpseg: Universal proposition learning for panoramic renal pathology segmentat...

    2024

  8. [16]

    , author Pan, Z

    author Ding, H. , author Pan, Z. , author Cen, Q. , author Li, Y. , author Chen, S. , year 2020 . title Multi-scale fully convolutional network for gland segmentation using three-class classification . journal Neurocomputing volume 380 , pages 150--161

    2020

  9. [17]

    , year 2020

    author Dosovitskiy, A. , year 2020 . title An image is worth 16x16 words: Transformers for image recognition at scale . journal arXiv preprint arXiv:2010.11929

    2020 arXiv

  10. [18]

    , author Litjens, G

    author Ehteshami Bejnordi, B. , author Litjens, G. , author Timofeeva, N. , author Otte-Höller, I. , author Homeyer, A. , author Karssemeijer, N. , author van der Laak, J.A. , year 2016 . title Stain specific standardization of whole-slide histopathological images . journal IE...

    2016

  11. [19]

    , author Swiderska-Chadaj, Z

    author Gallego, J. , author Swiderska-Chadaj, Z. , author Markiewicz, T. , author Yamashita, M. , author Gabaldon, M.A. , author Gertych, A. , year 2021 . title A u-net based framework to quantify glomerulosclerosis in digitized pas and h&e stained human tissues . journal Comp...

    2021

  12. [20]

    , author Lutnick, B

    author Ginley, B. , author Lutnick, B. , author Jen, K.Y. , author Fogo, A.B. , author Jain, S. , author Rosenberg, A. , author Walavalkar, V. , author Wilding, G. , author Tomaszewski, J.E. , author Yacoub, R. , et al., year 2019 . title Computational segmentation and classif...

    2019

  13. [21]

    , author Tomaszewski, J.E

    author Ginley, B. , author Tomaszewski, J.E. , author Sarder, P. , year 2017 . title Automatic computational labeling of glomerular textural boundaries , in: booktitle Medical Imaging 2017: Digital Pathology , organization SPIE . pp. pages 100--105

    2017

  14. [22]

    , author Kong, J

    author Gomes, J. , author Kong, J. , author Kurc, T. , author Melo, A.C. , author Ferreira, R. , author Saltz, J.H. , author Teodoro, G. , year 2021 . title Building robust pathology image analyses with uncertainty quantification . journal Computer Methods and Programs in Biom...

    2021

  15. [23]

    , author Lu, C.Z

    author Guo, M.H. , author Lu, C.Z. , author Hou, Q. , author Liu, Z. , author Cheng, M.M. , author Hu, S.M. , year 2022 . title Segnext: Rethinking convolutional attention design for semantic segmentation . journal Advances in Neural Information Processing Systems volume 35 , ...

    2022

  16. [24]

    , author Zhang, X

    author He, K. , author Zhang, X. , author Ren, S. , author Sun, J. , year 2016 . title Deep residual learning for image recognition , in: booktitle Proceedings of the IEEE conference on computer vision and pattern recognition , pp. pages 770--778

    2016

  17. [25]

    , author Mocanu, M

    author Hogan, J.J. , author Mocanu, M. , author Berns, J.S. , year 2016 . title The native kidney biopsy: update and evidence for best practice . journal Clinical Journal of the American Society of Nephrology volume 11 , pages 354--362

    2016

  18. [26]

    , author Shen, Y

    author Hu, E.J. , author Shen, Y. , author Wallis, P. , author Allen-Zhu, Z. , author Li, Y. , author Wang, S. , author Wang, L. , author Chen, W. , year 2021 . title Lora: Low-rank adaptation of large language models . journal arXiv preprint arXiv:2106.09685

    2021 arXiv

  19. [27]

    , author Lin, L

    author Huang, H. , author Lin, L. , author Tong, R. , author Hu, H. , author Zhang, Q. , author Iwamoto, Y. , author Han, X. , author Chen, Y.W. , author Wu, J. , year 2020 . title Unet 3+: A full-scale connected unet for medical image segmentation , in: booktitle ICASSP 2020-...

    2020

  20. [28]

    , author Carrero, J.J

    author Iglesias, P. , author Carrero, J.J. , author D \' ez, J.J. , year 2012 . title Gonadal dysfunction in men with chronic kidney disease: clinical features, prognostic implications and therapeutic options . journal Journal of nephrology volume 25 , pages 31

    2012

  21. [29]

    , author Jaeger, P.F

    author Isensee, F. , author Jaeger, P.F. , author Kohl, S.A. , author Petersen, J. , author Maier-Hein, K.H. , year 2021 . title nnu-net: a self-configuring method for deep learning-based biomedical image segmentation . journal Nature methods volume 18 , pages 203--211

    2021

  22. [30]

    , author Yang, H

    author Jha, A. , author Yang, H. , author Deng, R. , author Kapp, M.E. , author Fogo, A.B. , author Huo, Y. , year 2021 . title Instance segmentation for whole slide imaging: end-to-end or detect-then-segment . journal Journal of Medical Imaging volume 8 , pages 014001--014001

    2021

  23. [31]

    , author Chen, W

    author Jiang, L. , author Chen, W. , author Dong, B. , author Mei, K. , author Zhu, C. , author Liu, J. , author Cai, M. , author Yan, Y. , author Wang, G. , author Zuo, L. , author Shi, H. , year 2021 . title A deep learning-based approach for glomeruli instance segmentation ...

    2021 doi

  24. [32]

    , author Bajo, M.A

    author Jim \'e nez-Heffernan, J. , author Bajo, M.A. , author Perna, C. , author del Peso, G. , author Larrubia, J.R. , author Gamallo, C. , author S \'a nchez-Tomero, J. , author L \'o pez-Cabrera, M. , author Selgas, R. , year 2006 . title Mast cell quantification in normal ...

    2006

  25. [33]

    , year 2008

    author Kellum, J.A. , year 2008 . title Acute kidney injury . journal Critical care medicine volume 36 , pages S141--S145

    2008

  26. [34]

    , author Mintun, E

    author Kirillov, A. , author Mintun, E. , author Ravi, N. , author Mao, H. , author Rolland, C. , author Gustafson, L. , author Xiao, T. , author Whitehead, S. , author Berg, A.C. , author Lo, W.Y. , et al., year 2023 . title Segment anything , in: booktitle Proceedings of the...

    2023

  27. [35]

    , author Verma, R

    author Kumar, N. , author Verma, R. , author Sharma, S. , author Bhargava, S. , author Vahadane, A. , author Sethi, A. , year 2017 . title A dataset and a technique for generalized nuclear segmentation for computational pathology . journal IEEE transactions on medical imaging ...

    2017

  28. [36]

    , author Yang, H.C

    author Lim, B.J. , author Yang, H.C. , author Fogo, A.B. , year 2014 . title Animal models of regression/progression of kidney disease . journal Drug Discovery Today: Disease Models volume 11 , pages 45--51

    2014

  29. [37]

    , author Yang, J.W

    author Lim, B.J. , author Yang, J.W. , author Zou, J. , author Zhong, J. , author Matsusaka, T. , author Pastan, I. , author Zhang, M.Z. , author Harris, R.C. , author Yang, H.C. , author Fogo, A.B. , year 2017 . title Tubulointerstitial fibrosis can sensitize the kidney to su...

    2017

  30. [38]

    , author Doll \'a r, P

    author Lin, T.Y. , author Doll \'a r, P. , author Girshick, R. , author He, K. , author Hariharan, B. , author Belongie, S. , year 2017 . title Feature pyramid networks for object detection , in: booktitle Proceedings of the IEEE conference on computer vision and pattern recog...

    2017

  31. [39]

    , year 2024

    author Liu, Y. , year 2024 . title A hybrid cnn-transxnet approach for advanced glomerular segmentation in renal histology imaging . journal International Journal of Computational Intelligence Systems volume 17 , pages 126 . https://doi.org/10.1007/s44196-024-00523-7, :10.1007...

    2024 doi

  32. [40]

    , author Lin, Y

    author Liu, Z. , author Lin, Y. , author Cao, Y. , author Hu, H. , author Wei, Y. , author Zhang, Z. , author Lin, S. , author Guo, B. , year 2021 . title Swin transformer: Hierarchical vision transformer using shifted windows , in: booktitle Proceedings of the IEEE/CVF intern...

    2021

  33. [41]

    , author Mao, H

    author Liu, Z. , author Mao, H. , author Wu, C.Y. , author Feichtenhofer, C. , author Darrell, T. , author Xie, S. , year 2022 . title A convnet for the 2020s , in: booktitle Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pp. pages 11976--11986

    2022

  34. [42]

    , author Rodr \' guez-Ortega, A

    author Marti-Aguado, D. , author Rodr \' guez-Ortega, A. , author Mestre-Alagarda, C. , author Bauza, M. , author Valero-P \'e rez, E. , author Alfaro-Cervello, C. , author Benlloch, S. , author P \'e rez-Rojas, J. , author Ferr \'a ndez, A. , author Alemany-Monraval, P. , et ...

    2021

  35. [43]

    , author Lions, C

    author Mounier-Vehier, C. , author Lions, C. , author Devos, P. , author Jaboureck, O. , author Willoteaux, S. , author Carre, A. , author Beregi, J.P. , year 2002 . title Cortical thickness: an early morphological marker of atherosclerotic renal disease . journal Kidney inter...

    2002

  36. [44]

    , author Sadimin, E

    author Ren, J. , author Sadimin, E. , author Foran, D.J. , author Qi, X. , year 2017 . title Computer aided analysis of prostate histopathology images to support a refined gleason grading system , in: booktitle Medical Imaging 2017: Image Processing , organization Internationa...

    2017

  37. [45]

    , author Fischer, P

    author Ronneberger, O. , author Fischer, P. , author Brox, T. , year 2015 . title U-net: Convolutional networks for biomedical image segmentation , in: booktitle Medical image computing and computer-assisted intervention--MICCAI 2015: 18th international conference, Munich, Ger...

    2015

  38. [46]

    , author Chauhan, A

    author Singh Samant, S. , author Chauhan, A. , author DN, J. , author Singh, V. , year 2023 . title Glomerulus detection using segmentation neural networks . journal Journal of Digital Imaging volume 36 , pages 1633--1642 . https://doi.org/10.1007/s10278-022-00764-y, :10.1007/...

    2023 doi

  39. [47]

    , author He, Y

    author Tang, Y. , author He, Y. , author Nath, V. , author Guo, P. , author Deng, R. , author Yao, T. , author Liu, Q. , author Cui, C. , author Yin, M. , author Xu, Z. , et al., year 2024 . title Holohisto: End-to-end gigapixel wsi segmentation with 4k resolution sequential t...

    2024 arXiv

  40. [48]

    , author Reinke, A

    author Wiesenfarth, M. , author Reinke, A. , author Landman, B.A. , author Eisenmann, M. , author Saiz, L.A. , author Cardoso, M.J. , author Maier-Hein, L. , author Kopp-Schneider, A. , year 2021 . title Methods and open-source toolkit for analyzing and visualizing challenge r...

    2021

  41. [49]

    , author Moeckel, G

    author Wu, B. , author Moeckel, G. , year 2023 . title Application of digital pathology and machine learning in the liver, kidney and lung diseases . journal Journal of Pathology Informatics volume 14 , pages 100184 . https://www.sciencedirect.com/science/article/pii/S21533539...

    2023

  42. [50]

    , author Liu, Y

    author Xiao, T. , author Liu, Y. , author Zhou, B. , author Jiang, Y. , author Sun, J. , year 2018 . title Unified perceptual parsing for scene understanding , in: booktitle Proceedings of the European conference on computer vision (ECCV) , pp. pages 418--434

    2018

  43. [51]

    , author Wang, W

    author Xie, E. , author Wang, W. , author Yu, Z. , author Anandkumar, A. , author Alvarez, J.M. , author Luo, P. , year 2021 . title Segformer: Simple and efficient design for semantic segmentation with transformers . journal Advances in neural information processing systems v...

    2021

  44. [52]

    , author Lu, Y

    author Yao, T. , author Lu, Y. , author Long, J. , author Jha, A. , author Zhu, Z. , author Asad, Z. , author Yang, H. , author Fogo, A.B. , author Huo, Y. , year 2022 . title Glo-in-one: Holistic glomerular detection, segmentation, and lesion characterization with large-scale...

    2022 arXiv

  45. [53]

    , author Tang, Z

    author Yin, Y. , author Tang, Z. , author Weng, H. , year 2024 . title Application of visual transformer in renal image analysis . journal BioMed Eng OnLine volume 23 , pages 27 . https://doi.org/10.1186/s12938-024-01209-z, :10.1186/s12938-024-01209-z

    2024 doi

  46. [54]

    , author Yin, M

    author Yu, L. , author Yin, M. , author Deng, R. , author Liu, Q. , author Yao, T. , author Cui, C. , author Guo, J. , author Wang, Y. , author Wang, Y. , author Zhao, S. , author Yang, H. , author Huo, Y. , year 2024 . title Glo-in-one-v2: Holistic identification of glomerula...

    2024 arXiv

  47. [55]

    , author Yao, T

    author Yue, J. , author Yao, T. , author Deng, R. , author Liu, Q. , author Xiong, J. , author Guo, J. , author Yang, H. , author Huo, Y. , year 2024 a. title Weighted circle fusion: Ensembling circle representation from different object detection results . https://arxiv.org/a...

    2024 arXiv

  48. [56]

    , author Yao, T

    author Yue, J. , author Yao, T. , author Deng, R. , author Lu, S. , author Guo, J. , author Liu, Q. , author Yin, M. , author Xiong, J. , author Yang, H. , author Huo, Y. , year 2024 b. title Glofinder: Ai-empowered qupath plugin for wsi-level glomerular detection, visualizati...

    2024 arXiv

  49. [57]

    , author Han, D

    author Yun, S. , author Han, D. , author Oh, S.J. , author Chun, S. , author Choe, J. , author Yoo, Y. , year 2019 . title Cutmix: Regularization strategy to train strong classifiers with localizable features , in: booktitle Proceedings of the IEEE/CVF international conference...

    2019

  50. [58]

    , author Nan, Y

    author Zeng, C. , author Nan, Y. , author Xu, F. , author Lei, Q. , author Li, F. , author Chen, T. , author Liang, S. , author Hou, X. , author Lv, B. , author Liang, D. , et al., year 2020 . title Identification of glomerular lesions and intrinsic glomerular cell types in ki...

    2020

  51. [59]

    , year 2017

    author Zhang, H. , year 2017 . title mixup: Beyond empirical risk minimization . journal arXiv preprint arXiv:1710.09412

    2017 arXiv

  52. [60]

    , author Lu, J

    author Zheng, S. , author Lu, J. , author Zhao, H. , author Zhu, X. , author Luo, Z. , author Wang, Y. , author Fu, Y. , author Feng, J. , author Xiang, T. , author Torr, P.H. , et al., year 2021 . title Rethinking semantic segmentation from a sequence-to-sequence perspective ...

    2021

  53. [61]

    , author Rahman Siddiquee, M.M

    author Zhou, Z. , author Rahman Siddiquee, M.M. , author Tajbakhsh, N. , author Liang, J. , year 2018 . title Unet++: A nested u-net architecture for medical image segmentation , in: booktitle Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decisio...

    2018

  54. [62]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.