Pith. sign in

REVIEW 4 major objections 5 minor 58 references

MEATRD: Multimodal Anomalous Tissue Region Detection Enhanced with Spatial Transcriptomics

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read MEATRD claims that fusing histology images with spatial gene-expression profiles detects anomalous tissue regions that look visually normal, reporting an average AUC gain of 17.45 percentage points over nine baselines.

desk verdict F1 numbers in Table 1 are mutually inconsistent with the stated protocol; the AUC story may survive, but the F1 claims need correction before the paper's performance claims can be credited. read the letter →

arxiv 2412.10659 v1 pith:VOK5TQ2M submitted 2024-12-14 cs.CV cs.LGq-bio.QM

Kaichen Xu , Qilong Wu , Yan Lu , Yinan Zheng , Wenlin Li , Xingjie Tang , Jun Wang , Xiaobo Sun This is my paper · ORCID
classification cs.CVcs.LGq-bio.QM
keywords anomaloustissueregiondetectionspatialtranscriptomicshistologyimageanalysismultimodalanomalygraphattentiontransformerone-classclassificationreconstruction-basedinformationbottleneck
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

MEATRD is a proposed solution to a clinical blind spot: anomalous tissue regions that look almost identical to healthy tissue under the microscope escape image-only detectors. The paper argues that spatial transcriptomics can reveal such regions at the molecular level, and that fusing the two modalities through a graph transformer trained only on healthy reference tissue should catch them. The central claim is that this multimodal reconstruction-plus-one-class approach outperforms nine existing single-modality and multimodal anomaly-detection methods across eight breast cancer datasets, with a reported average AUC gain of 17.45 percentage points over the second-best method. The paper also claims a first theoretical result: the fused bottleneck encoding is a minimally sufficient representation of the modality-specific, task-relevant information, which underpins why the fusion works. If the claims hold, the method offers a label-free route to flagging visually subtle disease margins in tissue sections that already have gene-expression measurements.

What carries the argument

The central object is the masked graph dual-attention transformer (MGDAT), a graph transformer that treats each tissue spot as a node and fuses two modalities: a histology image patch embedding and a gene-expression embedding. Before a node is reconstructed, its own image and gene attributes are replaced with learnable mask tokens, forcing the network to rely on the k-nearest-neighbour spatial context and on cross-modal information from the other modality. Within each MGDAT block a low-dimensional fused bottleneck embedding $z_{fb}$ is produced by a transformer over the concatenated image and gene attributes, and this bottleneck is concatenated with the modality embeddings for graph attention message passing; the final image and gene embeddings decode into reconstructed patches and profiles. The second load-bearing mechanism is Stage III: latent encodings of original and reconstructed patches and profiles are compared to form a multimodal reconstruction error, and a one-class objective collapses inlier errors into a compact hypersphere, so inference scores each spot by distance from that center.

What would settle it

Recompute the reported F1 scores using a decision threshold chosen only from reference-data anomaly scores, with no knowledge of the true anomaly proportion, and independently re-annotate the target spots; if MEATRD's average F1 lead over the second-best baseline disappears, or its AUC on the re-annotated labels falls to within one standard deviation of the best image-only baseline, the central performance claim fails.

Watch

Extended reading notes

Core claim

MEATRD is presented as the first method that integrates histology images and spatial transcriptomics for anomalous tissue region detection. The paper's central claim is that a model trained only on healthy tissue can detect ATRs by reconstructing each spot's image patch and gene-expression profile from a fused multimodal graph embedding, and then measuring how far the latent reconstruction error lies from the healthy-spot center. The paper additionally claims a first theoretical analysis showing that the fused bottleneck encoding produced by MGDAT is a minimally sufficient representation of modality-specific, task-relevant information: it retains exactly the complementary signals needed to judge tissue normality and discards nuisance information. Empirically, the paper reports that MEATRD ranks first in AUC on all eight breast cancer datasets and first in F1 on six, with an average AUC gain of 17.45 percentage points over the second-best baseline, and that it detects a tumor edge region that visually resembles normal tissue.

Load-bearing premise

The headline numbers assume the spot-level pathology labels in the target datasets are accurate, complete, and consistent, the healthy reference datasets contain no anomalous tissue, and the F1 threshold is set with knowledge of the true anomaly proportion.

Editorial extensions

If this is right

  • ATRs that are visually indistinguishable from normal tissue can be detected from gene-expression differences, so tissue sections with ST data no longer need visible morphological change to be flagged.
  • Masking the target node's own information and collapsing latent reconstruction errors into a hypersphere gives a recipe that should reduce over-generalization in other reconstruction-based anomaly-detection settings.
  • The information-bottleneck result implies the fused encoding can be kept small without losing task-relevant signal; the paper finds 16 dimensions optimal, which lowers memory and compute.
  • Because MEATRD trains only on healthy reference tissue, it transfers to a new disease type (the paper reports PSC liver results) without requiring annotated anomalies for that disease.

Reading between the lines

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

  • A natural test the authors did not run: apply the same masked-bottleneck fusion to other paired image-plus-spatially-registered-data problems, such as satellite imagery with geochemical maps or whole-slide images with mass-spectrometry imaging.
  • The theory predicts that the optimal bottleneck dimension should track the amount of task-relevant shared information; measuring AUC against bottleneck size on datasets with different noise levels would test whether the 16-dimension optimum is a general law or a coincidence.
  • The F1 protocol uses the true anomaly proportion to set the threshold, which is unavailable at deployment; a threshold chosen from reference scores alone would be the more practical comparison, and the AUC numbers are the ones that would carry the argument.
  • If the bottleneck truly discards modality-specific nuisance information, MEATRD should be robust to stain variation in histology images; artificially altering stain appearance while keeping ST fixed would directly check that prediction.
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 / 5 minor

Summary. The paper proposes MEATRD, a multimodal method for anomalous tissue region (ATR) detection that combines histology image patches and spatial transcriptomics (ST) data. Tissue spots are modeled as nodes in a graph, and a masked graph dual-attention transformer (MGDAT) fuses image and gene embeddings into a low-dimensional bottleneck; the model is trained on normal reference tissue to reconstruct both modalities, and a one-class classifier is trained on latent multimodal reconstruction errors. The authors claim the method consistently ranks first in AUC and six times in F1 across eight breast cancer datasets, with an average AUC improvement of 17.45% over the second-best baseline, and that the fused bottleneck encoding is theoretically a minimally sufficient representation of modality-specific task-relevant information. The evaluation also includes four PSC liver datasets, ablation studies, sensitivity analyses, and a comparison on visually subtle tumor edge regions.

Significance. If the empirical results and the theoretical analysis are correct, the paper makes a useful contribution: it is among the first to integrate histology and ST for ATR detection, it provides a concrete architecture (MGDAT) for multimodal bottleneck fusion with node masking, and it releases code. The PSC generalization experiment and the careful documentation of preprocessing are strengths. However, the central empirical claim is currently undermined by internally inconsistent F1/AUC reporting, an unresolved inconsistency in the default weight parameter, and a model-selection protocol that appears to tune hyperparameters on the test sets. The information-theoretic result is a formal consequence of the information-bottleneck assumption rather than an empirical property of the implemented network, so the novelty claim should be stated more modestly.

major comments (4)
  1. [Experiments, Evaluation Protocols; Table 1] The F1 values reported in Table 1 cannot be reconciled with the stated evaluation protocol. The paper states that F1 is computed with the threshold matching the true anomaly proportion (Shenkar and Wolf 2021). On 10x-hBC-A1 (N=346, anomaly proportion 12.43%), selecting the top 43 spots by score yields an expected recall and precision of about 0.124 under an AUC of 0.5, giving F1 ≈ 0.124. The reported cell for scmap is AUC 0.500±0.000 and F1 0.934±0.000, which would require a near-perfect ranking of anomalies. This is impossible under the stated threshold protocol with a continuous score having AUC 0.5. The discrepancy invalidates the F1-based comparison in that table and, because the headline claim includes 'six times first in F1' and a 10.31% average F1 improvement, it undermines the F1 half of the central empirical claim. The authors should either correct the F1 computation, clarify the exact threshold rule, or remove the F1 claims until the numbers are reproducible.
  2. [Implementation Details; Sensitivity Analysis (Table 2); Discovering Anomalous Tissue Regions] The default value of β is reported inconsistently. The Implementation Details section states 'we have the three weight parameters α = 0.5 and β = 1', but Eq. (12)–(13) define β with 0 < β < 1, the main text describes decreasing β from 0.5 to 0.1, Table 2 marks β = 0.5 as the default, and the Robustness to Noisy Data section says 'we set β to 0.5'. This is not a minor typo: β controls the relative contribution of image versus gene modalities in Stage III, and the sensitivity analysis conclusions depend on which value is actually used. The manuscript must state one consistent default and ensure that the reported results, ablation, and sensitivity tables are generated with that value.
  3. [Experimental Settings; Sensitivity Analysis (Table 2)] The hyperparameter selection protocol is not specified, and it appears to be performed on the same test datasets used for the final performance table. Table 2 reports AUC and F1 averaged over the eight 10x-hBC test datasets for varying α, β, embedding dimensions, MGDAT layers, and attention heads; the defaults marked in gray are the best-performing settings on those datasets. No separate validation split or nested cross-validation is described. This creates a risk of selection bias in the reported headline numbers. The authors should clarify whether any of the eight test datasets were used to choose hyperparameters, and if so, report the model-selection procedure or provide a validation-based version of the sensitivity analysis.
  4. [Supplementary Material D, Assumption D.2 and Propositions D.1–D.2] The theoretical claim should be framed as a property of the information-bottleneck objective rather than of the MGDAT network itself. Assumption D.2 assumes that the encoders are information-lossless and that the fusion bottleneck follows the information-bottleneck objective; Propositions D.1 and D.2 then derive the optimum of that objective. The paper's wording that MEATRD's MGDAT generates a 'minimally sufficient representation' overstates what the proof establishes, since the proof does not verify that the trained network satisfies the assumed information-losslessness or the IB objective. At minimum, the contribution statement in the Introduction and Conclusion should be revised to say that the analysis characterizes the IB objective that motivates MGDAT, not that the trained model provably encodes exactly {b1, b2}.
minor comments (5)
  1. [Method, Masked Graph Dual-Attention Transformer] The acronym is inconsistently spelled 'MGADT' in the Method section; it should be 'MGDAT' throughout.
  2. [Algorithm 1] The notation in Algorithm 1 does not match Eq. (10): line 5 uses 'L_rec = L_ssim(P_b, P_hat_b) + λ L1(P_b, P_hat_b) + α L_SCE(X_b, X_hat_b)', while Eq. (10) defines the image loss as α·(−SSIM + L1) and the gene loss with weight (1−α). The algorithm should be aligned with the equations.
  3. [Supplement, Related Work] The reference to 'Shenkar and Wolf 2021' appears twice with different years (2021 and 2022) in the bibliography; the correct year should be used consistently.
  4. [Supplement, Ablation Studies] There is a typo 'Mobine-Unet' in the subsection title 'Mobine-Unet as pretrained visual feature extractor'; it should be 'Mobile-Unet'.
  5. [Table 1] scmap reports AUC 0.500±0.000 for every dataset; the authors should explain why this baseline has zero variance and why its F1 varies from 0.934 to 0.354 across datasets, as this pattern is consistent with a label-dependent or degenerate scoring behavior that needs clarification.

Circularity Check

1 steps flagged · score 6.0 of 10

Empirical comparison is external and non-circular; the claimed theoretical proof that MGDAT's bottleneck is minimally sufficient largely restates Assumption D.2's information-bottleneck objective.

  1. self definitional [Supplementary Material D, Assumption D.2 and Propositions D.1–D.2; main text Eq. (5) and the claim 'as theoretically demonstrated in supplementary material D']
    "Assuming f1 and f2 are information lossless encoders, and, along with the fusion bottleneck encoder f3, follow the information bottleneck theory proposed by Tishby et al., (Tishby, Pereira, and Bialek 2000). That is, z4 and z5 should be maximally informative about y with an information constrain on the bottleneck z3. ... Proposition D.2. Compactness of complementary task-relevant information. The objective functions in Assumption D.2 is minimized when: I(z3) = {b1, b2}"

    Assumption D.2 defines z3 as the solution of an information-bottleneck problem: maximize cM(z4; y) (and cM(z5; y)) subject to a bound on cM(z3; v1) (and cM(z3; v2)). The defining property of such a solution is that the bottleneck contains the task-relevant information transmissible under the constraint, which is exactly Proposition D.2's conclusion I(z3) = {b1,b2}; Proposition D.1 is the same trade-off read in the other direction. The main text's assertion that MGDAT's fused bottleneck provably 'collate[s] and condense[s] modality-specific, task-relevant information' therefore invokes its own supplementary proof, and the proof's load-bearing content is assumed in Assumption D.2 rather than derived from independent first principles.

full rationale

The empirical core of the paper is not circular: MEATRD is trained on healthy reference datasets (10x-hNB, 10x-hLiver) and evaluated on held-out disease datasets (10x-hBC, 10x-PSC) against nine external baselines, so the reported rankings are genuine out-of-sample evidence rather than a renaming of training data. I therefore do not treat the main performance claim as circular. The scmap F1 = 0.934 with AUC = 0.500 cell in Table 1 is internally inconsistent under the stated 'threshold matching the actual proportion of true anomalies' protocol; that is a correctness and reporting problem, not a circularity, and it should be corrected before the F1-based superiority claims are credited. The one substantive circular step is the theoretical section: Assumption D.2 defines the fused bottleneck as solving an information-bottleneck objective, and Propositions D.1–D.2 return the defining property of that objective, so the 'first theoretical analysis of the informational properties of multimodal bottleneck encoding' reduces by construction to its own assumption. The remaining self-citation (Xu et al. 2024 for the k = 6 neighbor choice in Definition D.2) is a minor implementation choice and not load-bearing. Overall, the central theoretical contribution carries partial circularity, while the empirical comparison remains self-contained against external data, giving a score of 6 rather than higher.

Assumptions & free parameters 7 free parameters · 3 assumptions · 0 invented entities

The empirical core uses no invented physical entities. The main hidden costs are the test-selected hyperparameters, the assumed lossless encoders in the theoretical section, and the assumption that pathological annotations are exhaustive and correct.

free parameters (7)
  • alpha (image reconstruction weight) = 0.5 (tuned)
    Weight between image and gene losses in Eq. (10); sensitivity Table 2a shows 0.5 best on the eight test datasets, so the value is chosen on the evaluation set.
  • beta (image weight in Stage III) = 0.5 in experiments; main text says 1
    Controls image versus gene contribution in Eqs. (12) and (13). The implementation section says beta=1 while sensitivity, Figure 3, and robustness discussion use 0.5 as the default.
  • bottleneck dimension D' = 16
    Fused bottleneck dimension in Eq. (5); sensitivity Table 2d chosen on the test data.
  • embedding dimension D = 256
    Visual and gene embedding dimension; sensitivity Table 2c chosen on the test data.
  • MGDAT layers and attention heads = 3 layers, 2 heads
    Sensitivity Tables 2f and 2g; values chosen on the test data.
  • gamma (SCE scaling factor) = not stated
    Scaling exponent in the gene reconstruction loss LSCE, Eq. (11); the value is not reported and affects the loss shape.
  • MAP-EM priors (a, b, kappa0, nu0) = a=1, b=10, kappa0=0.01, nu0=3
    Set heuristically from reference anomaly scores in supplementary Eqs. (25) and (26); the final anomaly threshold depends on these priors.
assumptions (3)
  • domain assumption Assumption D.1: histology image and ST are two views of the same biological information b, and the normality indicator y is determined by b with disjoint task-relevant components b0, b1, b2.
    Underlies the information-theoretic claim in supplementary D; not empirically validated in the paper.
  • ad hoc to paper Assumption D.2: image encoder f1 and gene encoder f2 are information-lossless, and the fusion bottleneck f3 follows the information bottleneck objective.
    The proof that MGDAT's bottleneck captures inclusive and compact information follows from this objective; the trained architecture is not shown to satisfy it.
  • domain assumption Reference healthy datasets contain no anomalous tissue and target datasets are exhaustively annotated.
    Required for computing AUC and F1 in Tables 1 and 7; any label noise biases the benchmark.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MEATRD: Multimodal Anomalous Tissue Region Detection Enhanced with Spatial Transcriptomics." pith.science (2026). https://pith.science/paper/VOK5TQ2M

@misc{pith2026241210659,
  author       = {Pith},
  title        = {Pith review of: MEATRD: Multimodal Anomalous Tissue Region Detection Enhanced with Spatial Transcriptomics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VOK5TQ2M}},
  note         = {Machine review of arXiv:2412.10659}
}
read the original abstract

The detection of anomalous tissue regions (ATRs) within affected tissues is crucial in clinical diagnosis and pathological studies. Conventional automated ATR detection methods, primarily based on histology images alone, falter in cases where ATRs and normal tissues have subtle visual differences. The recent spatial transcriptomics (ST) technology profiles gene expressions across tissue regions, offering a molecular perspective for detecting ATRs. However, there is a dearth of ATR detection methods that effectively harness complementary information from both histology images and ST. To address this gap, we propose MEATRD, a novel ATR detection method that integrates histology image and ST data. MEATRD is trained to reconstruct image patches and gene expression profiles of normal tissue spots (inliers) from their multimodal embeddings, followed by learning a one-class classification AD model based on latent multimodal reconstruction errors. This strategy harmonizes the strengths of reconstruction-based and one-class classification approaches. At the heart of MEATRD is an innovative masked graph dual-attention transformer (MGDAT) network, which not only facilitates cross-modality and cross-node information sharing but also addresses the model over-generalization issue commonly seen in reconstruction-based AD methods. Additionally, we demonstrate that modality-specific, task-relevant information is collated and condensed in multimodal bottleneck encoding generated in MGDAT, marking the first theoretical analysis of the informational properties of multimodal bottleneck encoding. Extensive evaluations across eight real ST datasets reveal MEATRD's superior performance in ATR detection, surpassing various state-of-the-art AD methods. Remarkably, MEATRD also proves adept at discerning ATRs that only show slight visual deviations from normal tissues.

Figures

Figures reproduced from arXiv: 2412.10659 by the authors.

Figure 1
Figure 1. Detecting ATRs with histology images and ST [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The workflow of MEATRD. masked before aggregating fused gene and imagery attributes of their neighboring nodes via attention-based mechanism. Formally speaking, let Gi(Vi , Ai , Zi) denote the subgraph of a target node i that covers up to its 3-hop neighbors, where Vi , Ai and Zi denote the node set, adjacency matrix, and node attribute matrix of Gi , respectively. We set the num￾ber of hops to be 3 as using more ho… view at source ↗
Figure 3
Figure 3. Visualized detection results of tumor edge regions that visually resemble the adjacent normal tissues in the 10x-hBC-I1 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Information diagrams of the two data modalities [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]
Figure 5
Figure 5. Figure 5: Violin plots illustrating the distributions of latent [PITH_FULL_IMAGE:figures/full_fig_p018_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 41 canonical work pages

  1. [1]

    Andersson, A.; Larsson, L.; Stenbeck, L.; et al. 2021. Spatial deconvolution of HER2-positive breast cancer delineates tumor-associated cell type interactions. Nat Commun, 12: 6012

    work page 2021

  2. [2]

    Arjovsky, M.; Chintala, S.; and Bottou, L. 2017. Wasserstein generative adversarial networks. In International conference on machine learning, 214--223. PMLR

    2017

  3. [3]

    E.; Veta, M.; Van Diest, P

    Bejnordi, B. E.; Veta, M.; Van Diest, P. J.; Van Ginneken, B.; Karssemeijer, N.; Litjens, G.; Van Der Laak, J. A.; Hermsen, M.; Manson, Q. F.; Balkenhol, M.; et al. 2017. Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. Jama, 318(22): 2199--2210

    work page 2017

  4. [4]

    Bergmann, P.; Fauser, M.; Sattlegger, D.; and Steger, C. 2019. MVTec AD -- A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    work page 2019

  5. [5]

    Cover, T. M. 1999. Elements of information theory. John Wiley & Sons

    1999

  6. [6]

    Deng, H.; and Li, X. 2022. Anomaly detection via reverse distillation from one-class embedding. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 9737--9746

    2022

  7. [7]

    Ding, K.; Li, J.; Bhanushali, R.; and Liu, H. 2019. Deep anomaly detection on attributed networks. In Proceedings of the 2019 SIAM International Conference on Data Mining, 594--602. SIAM

    work page 2019

  8. [8]

    a henb \

    Donahue, J.; Kr \"a henb \"u hl, P.; and Darrell, T. 2016. Adversarial feature learning. arXiv preprint arXiv:1605.09782

    arXiv 2016

Show all 58 references
  1. [9]

    Dong, K.; and Zhang, S. 2022. Deciphering spatial domains from spatially resolved transcriptomics with an adaptive graph attention auto-encoder. Nature communications, 13(1): 1739

    2022

  2. [10]

    He, K.; Fan, H.; Wu, Y.; Xie, S.; and Girshick, R. 2020. Momentum contrast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 9729--9738

    2020

  3. [11]

    He, K.; and Sun, J. 2015. Convolutional neural networks at constrained time cost. In Proceedings of the IEEE conference on computer vision and pattern recognition, 5353--5360

    2015

  4. [12]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, 770--778

    2016

  5. [13]

    Hou, Z.; He, Y.; Cen, Y.; Liu, X.; Dong, Y.; Kharlamov, E.; and Tang, J. 2023. GraphMAE2: A Decoding-Enhanced Masked Self-Supervised Graph Learner. In Proceedings of the ACM Web Conference 2023, 737--746

    2023

  6. [14]

    B.; Kadara, H.; Wang, L.; and Li, M

    Hu, J.; Coleman, K.; Zhang, D.; Lee, E. B.; Kadara, H.; Wang, L.; and Li, M. 2023. Deciphering tumor ecosystems at super resolution from spatial transcriptomics with TESLA. Cell systems, 14(5): 404--417

    2023

  7. [15]

    J.; Lee, E

    Hu, J.; Li, X.; Coleman, K.; Schroeder, A.; Ma, N.; Irwin, D. J.; Lee, E. B.; Shinohara, R. T.; and Li, M. 2021. SpaGCN: Integrating gene expression, spatial location and histology to identify spatial domains and spatially variable genes by graph convolutional network. Nature ...

    2021

  8. [16]

    N.; and Welling, M

    Kipf, T. N.; and Welling, M. 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907

    2016 arXiv

  9. [17]

    Y.; Yiu, A.; and Hemberg, M

    Kiselev, V. Y.; Yiu, A.; and Hemberg, M. 2018. scmap: projection of single-cell RNA-seq data across data sets. Nature methods, 15(5): 359--362

    2018

  10. [18]

    Komura, D.; and Ishikawa, S. 2018. Machine learning methods for histopathological image analysis. Computational and structural biotechnology journal, 16: 34--42

    2018

  11. [19]

    K.; Parra, E

    Kumar, G.; Pandurengan, R. K.; Parra, E. R.; Kannan, K.; and Haymaker, C. 2023. Spatial modelling of the tumor microenvironment from multiplex immunofluorescence images: methods and applications. Frontiers in immunology, 14: 1288802

    2023

  12. [20]

    Li, Z.; Wang, Y.; Ganan-Gomez, I.; Colla, S.; and Do, K.-A. 2022. A machine learning-based method for automatically identifying novel cells in annotating single-cell RNA-seq data. Bioinformatics, 38(21): 4885--4892

    2022

  13. [21]

    Liu, Z.; Zhou, Y.; Xu, Y.; and Wang, Z. 2023. Simplenet: A simple network for image anomaly detection and localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 20402--20411

    2023

  14. [22]

    Moses, L.; and Pachter, L. 2022. Museum of spatial transcriptomics. Nat Methods, 19: 534--546

    2022

  15. [23]

    Nagrani, A.; Yang, S.; Arnab, A.; Jansen, A.; Schmid, C.; and Sun, C. 2021. Attention bottlenecks for multimodal fusion. Advances in Neural Information Processing Systems, 34: 14200--14213

    2021

  16. [24]

    Okada, M.; and Taniguchi, T. 2021. Dreaming: Model-based reinforcement learning by latent imagination without reconstruction. In 2021 ieee international conference on robotics and automation (icra), 4209--4215. IEEE

    2021

  17. [25]

    P.; Chiang, M

    Ouardini, K.; Yang, H.; Unnikrishnan, B.; Romain, M.; Garcin, C.; Zenati, H.; Campbell, J. P.; Chiang, M. F.; Kalpathy-Cramer, J.; Chandrasekhar, V.; et al. 2019. Towards practical unsupervised anomaly detection on retinal images. In Domain Adaptation and Representation Transf...

    2019

  18. [26]

    Pan, J.; Liu, Y.; Zheng, Y.; and Pan, S. 2023. PREM: A Simple Yet Effective Approach for Node-Level Graph Anomaly Detection. arXiv preprint arXiv:2310.11676

    2023 arXiv

  19. [27]

    Paszke, A.; Gross, S.; Massa, F.; Lerer, A.; Bradbury, J.; Chanan, G.; Killeen, T.; Lin, Z.; Gimelshein, N.; Antiga, L.; et al. 2019. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing systems, 32

    2019

  20. [28]

    Pocevi c i \=u t \.e , M.; Eilertsen, G.; and Lundstr \"o m, C. 2021. Unsupervised anomaly detection in digital pathology using GANs. In 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI), 1878--1882. IEEE

    2021

  21. [29]

    Riasatian, A.; Babaie, M.; Maleki, D.; Kalra, S.; Valipour, M.; Hemati, S.; Zaveri, M.; Safarpoor, A.; Shafiei, S.; Afshari, M.; et al. 2021. Fine-tuning and training of densenet for histopathology image representation using tcga diagnostic slides. Medical Image Analysis, 70: 102032

    2021

  22. [30]

    T.; Nasrollahi, K.; Khan, F

    Ristea, N.-C.; Madan, N.; Ionescu, R. T.; Nasrollahi, K.; Khan, F. S.; Moeslund, T. B.; and Shah, M. 2022. Self-supervised predictive convolutional attentive block for anomaly detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 13576--13586

    2022

  23. [31]

    Roth, K.; Pemula, L.; Zepeda, J.; Sch \"o lkopf, B.; Brox, T.; and Gehler, P. 2022. Towards total recall in industrial anomaly detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 14318--14328

    2022

  24. [32]

    Rudolph, M.; Wehrbein, T.; Rosenhahn, B.; and Wandt, B. 2023. Asymmetric student-teacher networks for industrial anomaly detection. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2592--2602

    2023

  25. [33]

    A.; Binder, A.; M \"u ller, E.; and Kloft, M

    Ruff, L.; Vandermeulen, R.; Goernitz, N.; Deecke, L.; Siddiqui, S. A.; Binder, A.; M \"u ller, E.; and Kloft, M. 2018. Deep one-class classification. In International conference on Machine Learning, 4393--4402. PMLR

    2018

  26. [34]

    Sabour, S.; Frosst, N.; and Hinton, G. E. 2017. Dynamic routing between capsules. Advances in neural information processing systems, 30

    2017

  27. [35]

    M.; Langs, G.; and Schmidt-Erfurth, U

    Schlegl, T.; Seeb \"o ck, P.; Waldstein, S. M.; Langs, G.; and Schmidt-Erfurth, U. 2019. f-AnoGAN: Fast unsupervised anomaly detection with generative adversarial networks. Medical image analysis, 54: 30--44

    2019

  28. [36]

    Shen, R.; Liu, L.; Wu, Z.; Zhang, Y.; Yuan, Z.; Guo, J.; Yang, F.; Zhang, C.; Chen, B.; Feng, W.; et al. 2022. Spatial-ID: a cell typing method for spatially resolved transcriptomics via transfer learning and spatial embedding. Nature communications, 13(1): 7640

    2022

  29. [37]

    Shenkar, T.; and Wolf, L. 2021. Anomaly detection for tabular data with internal contrastive learning. In International Conference on Learning Representations

    2021

  30. [38]

    Shenkar, T.; and Wolf, L. 2022. Anomaly detection for tabular data with internal contrastive learning. In International conference on learning representations

    2022

  31. [39]

    Shvetsova, N.; Bakker, B.; Fedulova, I.; Schulz, H.; and Dylov, D. V. 2021. Anomaly detection in medical imaging with deep perceptual autoencoders. IEEE Access, 9: 118571--118583

    2021

  32. [40]

    Simonyan, K.; and Zisserman, A. 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556

    2014 arXiv

  33. [41]

    Sohn, K.; Li, C.-L.; Yoon, J.; Jin, M.; and Pfister, T. 2020. Learning and evaluating representations for deep one-class classification. arXiv preprint arXiv:2011.02578

    2020 arXiv

  34. [42]

    L.; Ciga, O.; and Martel, A

    Srinidhi, C. L.; Ciga, O.; and Martel, A. L. 2021. Deep neural network models for computational histopathology: A survey. Medical image analysis, 67: 101813

    2021

  35. [43]

    Tian, Y.; Sun, C.; Poole, B.; Krishnan, D.; Schmid, C.; and Isola, P. 2020. What makes for good views for contrastive learning? Advances in neural information processing systems, 33: 6827--6839

    2020

  36. [44]

    C.; and Bialek, W

    Tishby, N.; Pereira, F. C.; and Bialek, W. 2000. The information bottleneck method. arXiv preprint physics/0004057

    2000 arXiv

  37. [45]

    E.; and Gadermayr, M

    Tschuchnig, M. E.; and Gadermayr, M. 2022. Anomaly detection in medical imaging-a mini review. In Data Science--Analytics and Applications: Proceedings of the 4th International Data Science Conference--iDSC2021, 33--38. Springer

    2022

  38. [46]

    Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Li \`o , P.; and Bengio, Y. 2018. Graph Attention Networks. In International Conference on Learning Representations

    2018

  39. [47]

    Wang, G.; Wu, S.; Xiong, Z.; Qu, H.; Fang, X.; and Bao, Y. 2024. CROST: a comprehensive repository of spatial transcriptomics. Nucleic Acids Research, 52(D1): D882--D890

    2024

  40. [48]

    Wang, L.; Maletic-Savatic, M.; and Liu, Z. 2022. Region-specific denoising identifies spatial co-expression patterns and intra-tissue heterogeneity in spatially resolved transcriptomics data. Nature Communications, 13(1): 6912

    2022

  41. [49]

    Wang, Y.; Peng, J.; Zhang, J.; Yi, R.; Wang, Y.; and Wang, C. 2023. Multimodal Industrial Anomaly Detection via Hybrid Fusion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8032--8041

    2023

  42. [50]

    Wang, Y.; Song, B.; Wang, S.; Chen, M.; Xie, Y.; Xiao, G.; Wang, L.; and Wang, T. 2022. Sprod for de-noising spatially resolved transcriptomics data based on position and image information. Nature methods, 19(8): 950--958

    2022

  43. [51]

    A.; Angerer, P.; and Theis, F

    Wolf, F. A.; Angerer, P.; and Theis, F. J. 2018. SCANPY: large-scale single-cell gene expression data analysis. Genome biology, 19: 1--5

    2018

  44. [52]

    Xu, K.; Lu, Y.; Hou, S.; Liu, K.; Du, Y.; Huang, M.; Feng, H.; Wu, H.; and Sun, X. 2024. Detecting anomalous anatomic regions in spatial transcriptomics with STANDS. Nature Communications, 15(1): 8223

    2024

  45. [53]

    Yi, J.; and Yoon, S. 2020. Patch svdd: Patch-level svdd for anomaly detection and segmentation. In Proceedings of the Asian conference on computer vision

    2020

  46. [54]

    N.; Sullivan, R.; and Hu, F

    Zehnder, P.; Feng, J.; Fuji, R. N.; Sullivan, R.; and Hu, F. 2022. Multiscale generative model using regularized skip-connections and perceptual loss for anomaly detection in toxicologic histopathology. Journal of Pathology Informatics, 13: 100102

    2022

  47. [55]

    Zingman, I.; Stierstorfer, B.; Lempp, C.; and Heinemann, F. 2023. Learning image representations for anomaly detection: application to discovery of histological alterations in drug development. Medical Image Analysis, 103067

    2023

  48. [56]

    Zong, Y.; Yu, T.; Wang, X.; Wang, Y.; Hu, Z.; and Li, Y. 2022. conST: an interpretable multi-modal contrastive learning framework for spatial transcriptomics. bioRxiv, 2022--01

    2022

  49. [57]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  50. [58]

    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 11, 2026 · model on record in the stance chip above.