Bayesian Autoencoder for Medical Anomaly Detection: Uncertainty-Aware Approach for Brain 2 MRI Analysis

Dip Roy

REVIEW 5 major objections 5 minor 25 references

Bayesian Autoencoder for Medical Anomaly Detection: Uncertainty-Aware Approach for Brain 2 MRI Analysis

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A Bayesian VAE with attention detects brain MRI anomalies at ROC AUC 0.834.

desk verdict Incremental VAE architecture with a sensible uncertainty-weighted score, but the tiny test set and inconsistent reported metrics sink the headline claim that uncertainty is essential. read the letter →

arxiv 2504.15562 v1 pith:YVELUTH7 submitted 2025-04-22 cs.LG cs.CV

Dip Roy This is my paper · ORCID

classification cs.LGcs.CV

keywords MedicalimagingAnomalydetectionBayesianneuralnetworksVariationalautoencoderUncertaintyestimationAttentionmechanismsBrainMRI

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that a variational autoencoder trained on mostly normal brain MRI slices can flag anomalous scans by weighting its reconstruction error with estimated uncertainty. It introduces a Bayesian VAE with multi-head attention that outputs both a mean reconstruction and per-pixel variances, and averages several latent-space reconstructions at test time to form an uncertainty-aware anomaly score. On the BraTS2020 T1ce data the model reaches ROC AUC 0.834 and PR AUC 0.833, and ablation experiments report that removing the attention mechanism or either uncertainty component lowers performance. The payoff, if the claim holds, is an anomaly detector that gives clinicians not just a flag but a confidence estimate tied to the reconstruction.

What carries the argument

The load-bearing object is the uncertainty-weighted reconstruction score built from a Bayesian VAE with multi-head attention. The encoder compresses each image to latent mean and log-variance parameters; the decoder emits both a reconstructed image and a per-pixel log-variance, so the training objective is a $\beta$-weighted ELBO with a Gaussian negative log-likelihood reconstruction term. At inference, $K$ latent samples produce $K$ reconstructions: their mean gives $\bar{x}$, their variance gives the term the paper calls epistemic uncertainty, and the average of the decoder variances gives the aleatoric term. Multi-head scaled dot-product attention after selected convolutional layers refines the feature maps, and the final score $A_{\mathrm{pixel}}(x)=(x-\bar{x})^2/(U_{\mathrm{total}}+\epsilon)$ is what converts reconstruction residuals into an anomaly map that explicitly suppresses high-uncertainty regions.

What would settle it

Train the exact same architecture and training procedure but at inference use a single latent sample ($K=1$) instead of $K=5$, and compare ROC AUC on the same test set. If the single-sample score matches or exceeds the multi-sample score, the sampling-based uncertainty term is not what carries detection performance.

Watch

Extended reading notes

Core claim

The paper's central claim is that uncertainty estimation is a necessary ingredient for reconstruction-based anomaly detection in brain MRI, not a post-hoc add-on. Concretely, the pixel-level score is $A_{\mathrm{pixel}}(x)=(x-\bar{x})^2/(U_{\mathrm{total}}+\epsilon)$, where $\bar{x}$ is the mean of $K$ reconstructions sampled from the latent code and $U_{\mathrm{total}}$ is the sum of the variance across those reconstructions and the decoder's predicted per-pixel variance. With $K=5$, the model assigns abnormal slices an average anomaly score of 0.0127 versus 0.0058 for normal slices, giving ROC AUC 0.834 and PR AUC 0.833 on BraTS2020. The ablation results, which drop multi-head attention, aleatoric uncertainty, or epistemic uncertainty and observe lower AUC in every case, are offered as evidence that both uncertainty components carry complementary signal.

Load-bearing premise

The load-bearing premise is that the spread of reconstructions from re-sampling the latent code measures the model's uncertainty about its parameters; since the VAE's weights are deterministic after training, that spread measures sensitivity to the latent draw, and if that identification fails the uncertainty decomposition and the uncertainty-weighted score lose their stated meaning.

Editorial extensions

If this is right

If the central claim is correct, an unsupervised model trained on normal slices can produce image-level anomaly scores that separate abnormal from normal brain MRI slices at ROC AUC 0.834, with an optimal threshold near 0.097 and F1 of 0.818.
The ablation pattern implies that multi-head attention, aleatoric uncertainty, and the sampling-based uncertainty term are each load-bearing: removing any one drops ROC AUC from 0.834 to between 0.790 and 0.801, and dropping both attention and Bayesian terms leaves a deterministic autoencoder at 0.751.
Because total uncertainty used directly as a discriminator reaches ROC AUC 0.846 while the combined anomaly score reaches 0.834, the uncertainty maps themselves carry much of the separation signal.
At inference the model draws $K=5$ latent samples and averages pixel-wise variances, so the reported performance comes with a tunable computational cost: more samples buy tighter variance estimates at proportionally higher inference time.

Reading between the lines

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

A stricter test of the epistemic-uncertainty claim would vary $K$ (for example, 1 versus 5 versus 20) and show that AUC improves with $K$; the paper fixes $K=5$ and ablates whole components, so it cannot distinguish the sampling mechanism from the architecture's overall capacity.
The paper's 'epistemic' variance is probably better read as sensitivity to the latent draw than as Bayesian weight uncertainty; a reader who needs parameter uncertainty should look for explicit weight distributions or dropout-based sampling rather than the variance formula in Section 3.4.
The same uncertainty-weighted score could be applied to 3D volumes or multiple MRI modalities; the paper lists these as future work, and a volumetric extension would likely change the uncertainty maps because slice-level reconstruction ignores through-plane context.
One could turn the qualitative boundary-localization result into a quantitative test by thresholding the uncertainty map and measuring overlap with tumor segmentation masks; the paper does not report such a metric.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Colleague,

Quick read of Roy's Bayesian VAE plus multi-head attention for brain MRI anomaly detection. The actual novelty is modest but real: applying multi-head attention in both encoder and decoder of a VAE and feeding the result into an uncertainty-weighted anomaly score (Eqs 14–15) is not in the papers it cites. The writing is clear, the method section is specific enough to reimplement, and the qualitative uncertainty maps around tumor boundaries are a plausible clinical aid.

The problem is the evaluation. The entire quantitative case rests on 60 test slices (30 normal, 30 abnormal). With n=60, the standard error of an AUC near 0.834 is roughly 0.048, so the ablation gaps in Table 3 — 0.039 ROC without attention, 0.023 PR without aleatoric — are within one standard error. No confidence intervals, no repeated runs, no code. Worse, the same full model is reported with PR AUC 0.833 in Table 1, 0.805 in Table 3, and 0.805 in the Figure 1 caption. That internal inconsistency alone makes the exact numbers unreliable. The claim that uncertainty is an 'essential component' is not supported by the evidence as presented.

The conceptual issue with Eq. 11 is real but secondary. The variance across K latent-code samples measures sensitivity to the latent draw, not Bayesian parameter uncertainty, since the VAE weights are deterministic. The paper should call it something like 'latent sampling variance' or justify a proper posterior over weights. Even with the rename, the current test size still can't distinguish the ablations from noise.

It is a fair paper in the sense that it clearly states its limitations (2D, single modality, computational cost), and the core idea of uncertainty-weighted scoring is reasonable. But the reported numbers are not reproducible from the manuscript alone, and the central empirical claim needs a larger test set, calibrated ablations, and at least a consistency check across metrics.

Should it go to peer review? Yes — not because the current results are convincing, but because the architecture and scoring rule are worth a proper test. A serious referee should ask for a real evaluation. As it stands, I would not cite the reported AUCs, and I would not bring it to reading group unless the discussion is about evaluation pitfalls.

Recommendation: send to review, with the expectation that the empirical section gets rebuilt.

Referee Report

5 major / 5 minor

Summary. The paper proposes a variational autoencoder with multi-head attention for unsupervised anomaly detection in brain MRI (BraTS2020 T1ce slices). During inference, the model draws K latent samples, computes mean reconstruction and two uncertainty terms: an 'epistemic' term from sample variance of reconstructions and an 'aleatoric' term from the decoder's predicted per-pixel variance. These are combined into a total uncertainty map and an anomaly score with an additional weighting parameter alpha. The authors report ROC AUC 0.834 and PR AUC 0.833, compare against uncertainty-aware baselines, present ablations removing attention and each uncertainty term, and show qualitative uncertainty maps. They conclude that uncertainty modeling and attention are essential components of their anomaly detection performance.

Significance. If the empirical claims were statistically robust, the paper would make a useful contribution: the idea of folding uncertainty estimates into the anomaly score is clinically motivated, and the qualitative observation that uncertainty concentrates at tumor boundaries is interesting. The paper also provides a clear architectural template that others could build on. However, the current evidence base is too weak to support the strong conclusion that uncertainty modeling is 'essential': the test set contains only 60 slices, the reported PR AUC for the full model is internally inconsistent, and the main ablation gaps are within the noise level of such a small evaluation. The conceptual identification of latent-sample variance with epistemic model uncertainty is also not justified. These issues are fixable, but they require substantially stronger evaluation and a corrected interpretation of the uncertainty terms.

major comments (5)

[Section 5.1, Table 1 vs Table 3 and Fig. 1] The full model is assigned three different PR AUC values: 0.833 in Table 1, 0.805 in Table 3, and 0.805 in the Figure 1 caption. This internal inconsistency means the headline performance number is not reproducible from the manuscript as written. Please correct the values and explain the discrepancy.
[Section 4.1 and Section 5.4] The test set consists of only 60 balanced slices (30 normal, 30 abnormal), and no confidence intervals or repeated-seed results are reported. For an AUC near 0.834 with n=60, the standard error is on the order of 0.05, so the ablation differences in Table 3 (e.g., ROC AUC drop of 0.039 when attention is removed, PR AUC drop of 0.023 when aleatoric uncertainty is removed) are within one standard error. The claim that either uncertainty component or attention is 'essential' is therefore not statistically supported by the presented evidence.
[Section 3.4, Eq. (11)] Equation (11) labels the variance across K latent-space samples as 'epistemic uncertainty', but in a standard VAE the network weights are deterministic after training; sampling the latent code z does not sample model parameters. This quantity measures sensitivity to the latent draw, not Bayesian model-parameter uncertainty. The paper either needs to justify this identification, or rename the term (e.g., 'latent sampling uncertainty') and revise the claim that the model captures both epistemic and aleatoric uncertainty. As written, the uncertainty decomposition and the uncertainty-weighted anomaly score rest on a mislabeled quantity.
[Section 3.4, Eq. (15), and Section 4.2] The image-level anomaly score A(x) depends on an unreported weighting parameter alpha, and Figure 1 reports an F1 value computed at a threshold selected from the test ROC curve. The value of alpha, the threshold-selection procedure, and whether the threshold was chosen on a validation set rather than the test set must be reported. Otherwise the operating point (F1 = 0.818) and any conclusions drawn from it include an element of test-set fitting.
[Section 4.1] The manuscript does not state whether the 60 test slices come from the same patients as the training slices, nor how many patient volumes are represented. If slices from the same patient appear in both training and test, the reported AUCs may be optimistically biased. Patient-level or volume-level splitting should be described and, if possible, used for evaluation.

minor comments (5)

[Figure 1] The caption states 'F1 = 0.818' while the abstract and Table 1 report PR AUC 0.833; please clarify whether these are different quantities and where the F1 threshold came from.
[Section 3.4, Eq. (14)] The anomaly score divides the squared error by total uncertainty, which explicitly lowers the score in high-uncertainty regions; since abnormal regions may themselves exhibit high uncertainty (as shown in the qualitative results), the rationale for this weighting should be discussed more carefully.
[Section 5.3, Table 2] The comparison with Probabilistic U-Net, MC-Dropout CNN, and Deep Ensemble reports only point estimates with no confidence intervals or description of how these baselines were configured and trained; the comparison would be more informative with error bars and implementation details.
[Figure 4] The caption says 'Training dynamics showing total uncertainty across epochs' but the text does not explain how total uncertainty is aggregated over the dataset or which split this curve comes from.
[Throughout] The term 'Bayesian VAE' is potentially misleading because the model does not perform Bayesian inference over network weights; consider using 'variational autoencoder with latent-space sampling' or clearly state that the weights are deterministic.

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation circularity; the only construction-based issue is a test-set-optimized F1 threshold, plus an unreported alpha that weakens reproducibility.

other [Figure 1 caption (Section 5, Results)]
"Fig. 1 ROC Curve (Optimal threshold = 0.097) and Precision-Recall Curve (F1 = 0.818). The ROC curve shows an AUC of 0.834, while the PR curve shows an AUC of 0.805."

The caption identifies a threshold of 0.097 as 'optimal' and reports F1 = 0.818 at that threshold. If the threshold is selected by scanning the test ROC/PR curve, then the reported F1 is computed at a threshold that maximizes (or nearly maximizes) test performance, so this particular F1 value is fitted to the test set by construction rather than being an independent, pre-specified operating-point estimate. This is a mild fitted-input/reported-metric coupling; it does not affect the derivation of the anomaly score or the main ROC/PR AUC claims.

full rationale

The paper's derivation chain is self-contained and not circular: the VAE ELBO (Eqs. 7-9), latent sampling (Eq. 5), reconstruction and uncertainty estimates (Eqs. 10-13), and the anomaly score (Eqs. 14-15) are all defined from the model's own outputs and standard variational objectives, not from anomaly labels or from the claimed result. The central claim that uncertainty modeling is essential rests on ablations (Table 3); those ablations are empirical comparisons and their statistical fragility (n=60, no confidence intervals, inconsistent PR AUC values) is a correctness/robustness concern, not circularity. Equation 11 labels latent-draw variance as 'epistemic uncertainty,' which is conceptually questionable because the VAE weights are deterministic, but that is a validity/terminology issue rather than a circular reduction. The only concrete construction-based issue is the 'Optimal threshold' F1 in Figure 1, where the threshold appears to be selected from the test curve and the F1 is then reported from that same fitted threshold. Additionally, Eq. 15 depends on an unreported alpha, which is a transparency/reproducibility concern, but the paper gives no evidence that alpha was tuned on the test set, so it cannot be counted as a demonstrated circular step. Overall, no significant derivation circularity; score 2 reflects the minor test-set-fitted F1 reporting and the unstated alpha.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities. Its main free parameters are standard hyperparameters and an unreported alpha; the key ad hoc assumption is that latent sampling yields epistemic uncertainty.

free parameters (4)

beta (ELBO weight) = 0.1
Set to 0.1 in Section 4.2, no sensitivity analysis; affects the latent space regularity and reconstruction balance.
alpha (anomaly score weight)
Introduced in Eq 15, no value or tuning procedure reported; if tuned on the test set it would be a free parameter affecting reported AUC.
K (latent samples at inference) = 5
Used to compute epistemic uncertainty in Eq 11; no analysis of sensitivity to K; small K makes the variance estimate noisy.
decision threshold = 0.097
Optimal threshold selected from the test ROC curve to report F1=0.818; fitting to the test set inflates the reported F1.

assumptions (6)

standard math VAE ELBO and reparameterization trick are valid and used as stated
Equations 5-9 rely on standard VAE theory; no derivation issues.
standard math Scaled dot-product multi-head attention is implemented as in Eqs 2-4
Attention mechanism is a standard building block.
domain assumption The training set is predominantly normal and representative of the normal brain MRI distribution
Section 4.1 states 85% of normal slices are used for training, but the selection of abnormal slices and the balance assumption are not fully specified.
domain assumption The BraTS2020 tumor segmentation masks provide a reliable normal/abnormal label for each slice
Slices are categorized as normal or abnormal based on tumor presence; label noise or mask errors would affect the evaluation.
ad hoc to paper Variance across K latent samples estimates epistemic model uncertainty
Section 3.4, Eq 11 equates this variance to epistemic uncertainty without theoretical justification; standard Bayesian deep learning estimates model uncertainty via weight posteriors or MC dropout.
ad hoc to paper The uncertainty-weighted anomaly score in Eqs 14-15 is an appropriate score
The division by total uncertainty and the alpha-weighted combination are introduced without derivation or external validation.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Bayesian Autoencoder for Medical Anomaly Detection: Uncertainty-Aware Approach for Brain 2 MRI Analysis." pith.science (2026). https://pith.science/paper/YVELUTH7

@misc{pith2026250415562,
  author       = {Pith},
  title        = {Pith review of: Bayesian Autoencoder for Medical Anomaly Detection: Uncertainty-Aware Approach for Brain 2 MRI Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YVELUTH7}},
  note         = {Machine review of arXiv:2504.15562}
}

read the original abstract

In medical imaging, anomaly detection is a vital element of healthcare diagnostics, especially for neurological conditions which can be life-threatening. Conventional deterministic methods often fall short when it comes to capturing the inherent uncertainty of anomaly detection tasks. This paper introduces a Bayesian Variational Autoencoder (VAE) equipped with multi-head attention mechanisms for detecting anomalies in brain magnetic resonance imaging (MRI). For the purpose of improving anomaly detection performance, we incorporate both epistemic and aleatoric uncertainty estimation through Bayesian inference. The model was tested on the BraTS2020 dataset, and the findings were a 0.83 ROC AUC and a 0.83 PR AUC. The data in our paper suggests that modeling uncertainty is an essential component of anomaly detection, enhancing both performance and interpretability and providing confidence estimates, as well as anomaly predictions, for clinicians to leverage in making medical decisions.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 20 canonical work pages

[1]

Journal of Medical Systems 45(4), 1–16 (2021)

Alloqmani, Y., et al.: Deep learning approaches for detection and diagnosis of brain tumors: A survey. Journal of Medical Systems 45(4), 1–16 (2021)

work page 2021
[2]

IEEE Access 9, 121156–121183 (2021)

Masood, A., et al.: A comprehensive survey of brain tumor detection in MRI: A critical review. IEEE Access 9, 121156–121183 (2021)

work page 2021
[3]

Artificial Intelligence in Medicine 121, 102197 (2021)

Fernando, K.R.M., et al.: Deep learning for biomedical image reconstruction: A survey. Artificial Intelligence in Medicine 121, 102197 (2021)

work page 2021
[4]

Medical Image Analysis 54, 10–19 (2019)

Zhang, J., et al.: Medical image classification using synergic deep learning. Medical Image Analysis 54, 10–19 (2019)

work page 2019
[5]

ACM Computing Surveys 41(3), 1–58 (2009)

Chandola, V., Banerjee, A., Kumar, V.: Anomaly detection: A survey. ACM Computing Surveys 41(3), 1–58 (2009)

work page 2009
[6]

Medical Image Analysis 8(3), 275–283 (2004)

Prastawa, M., Bullitt, E., Ho, S., Gerig, G.: A brain tumor segmentation framework based on outlier detection. Medical Image Analysis 8(3), 275–283 (2004)

work page 2004
[7]

Special Lecture on IE 2(1), 1–18 (2015)

An, J., Cho, S.: Variational autoencoder based anomaly detection using recon- struction probability. Special Lecture on IE 2(1), 1–18 (2015)

work page 2015
[8]

arXiv preprint arXiv:1812.05941 (2019)

Zimmerer, D., Kohl, S.A., Petersen, J., Isensee, F., Maier-Hein, K.H.: Context- encoding variational autoencoder for unsupervised anomaly detection. arXiv preprint arXiv:1812.05941 (2019)

arXiv 2019

Show all 25 references

[9]

In: International Conference on Information Processing in Medical Imaging, pp

Schlegl, T., Seeb¨ ock, P., Waldstein, S.M., Schmidt-Erfurth, U., Langs, G.: Unsu- pervised anomaly detection with generative adversarial networks to guide marker discovery. In: International Conference on Information Processing in Medical Imaging, pp. 146–157 (2017) 14

2017
[10]

In: International MICCAI Brainlesion Workshop, pp

Baur, C., Wiestler, B., Albarqouni, S., Navab, N.: Deep autoencoding models for unsupervised anomaly segmentation in brain MR images. In: International MICCAI Brainlesion Workshop, pp. 161–169 (2018)

2018
[11]

arXiv preprint arXiv:1806.04972 (2018)

Chen, X., Konukoglu, E.: Unsupervised detection of lesions in brain MRI using constrained adversarial auto-encoders. arXiv preprint arXiv:1806.04972 (2018)

2018 arXiv
[12]

Nature Machine Intelligence 1(1), 20–23 (2019)

Begoli, E., Bhattacharya, T., Kusnezov, D.: The need for uncertainty quantifi- cation in machine-assisted medical decision making. Nature Machine Intelligence 1(1), 20–23 (2019)

2019
[13]

Scientific Reports 7(1), 1–14 (2017)

Leibig, C., Allken, V., Ayhan, M.S., Berens, P., Wahl, S.: Leveraging uncertainty information from deep neural networks for disease detection. Scientific Reports 7(1), 1–14 (2017)

2017
[14]

Neal, R.M.: Bayesian learning for neural networks. vol. 118, Springer Science & Business Media (2012)

2012
[15]

5574–5584 (2017)

Kendall, A., Gal, Y.: What uncertainties do we need in bayesian deep learning for computer vision? In: Advances in Neural Information Processing Systems, pp. 5574–5584 (2017)

2017
[16]

In: Medical Imaging with Deep Learning Conference (2018)

Pawlowski, N., Lee, M.C., Rajchl, M., McDonagh, S., Ferrante, E., Kamnitsas, K., Glocker, B.: Unsupervised lesion detection in brain CT using bayesian con- volutional autoencoders. In: Medical Imaging with Deep Learning Conference (2018)

2018
[17]

In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp

Abati, D., Porrello, A., Calderara, S., Cucchiara, R.: Latent space autoregression for novelty detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 481–490 (2019)

2019
[18]

In: International Conference on Machine Learning, pp

Gal, Y., Ghahramani, Z.: Dropout as a bayesian approximation: Representing model uncertainty in deep learning. In: International Conference on Machine Learning, pp. 1050–1059 (2016)

2016
[19]

Medical Image Analysis 53, 197–207 (2019)

Schlemper, J., Oktay, O., Schaap, M., Heinrich, M., Kainz, B., Glocker, B., Rueck- ert, D.: Attention gated networks: Learning to leverage salient regions in medical images. Medical Image Analysis 53, 197–207 (2019)

2019
[20]

arXiv preprint arXiv:1801.09927 (2018)

Guan, Q., Huang, Y., Zhong, Z., Zheng, Z., Zheng, L., Yang, Y.: Diagnose like a radiologist: Attention guided convolutional neural network for thorax disease classification. arXiv preprint arXiv:1801.09927 (2018)

2018 arXiv
[21]

In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp

Wang, X., Peng, Y., Lu, L., Lu, Z., Bagheri, M., Summers, R.M.: Chestx-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classi- fication and localization of common thorax diseases. In: Proceedings of the IEEE Conference on Computer Vision and Patte...

2017
[22]

arXiv preprint arXiv:1804.03999 (2018)

Oktay, O., Schlemper, J., Folgoc, L.L., Lee, M., Heinrich, M., Misawa, K., Glocker, B.: Attention u-net: Learning where to look for the pancreas. arXiv preprint arXiv:1804.03999 (2018)

2018 arXiv
[23]

In: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, pp

Guo, Y., Yuan, H.: Anomaly detection in industrial images using attention pyra- mid network. In: Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, pp. 4177–4181 (2020)

2020
[24]

In: Advances in Neural Information Processing Systems, pp

Kohl, S.A., Romera-Paredes, B., Meyer, C., De Fauw, J., Ledsam, J.R., Maier- Hein, K.H., Eslami, S.M., Rezende, D.J., Ronneberger, O.: A probabilistic U- Net for segmentation of ambiguous images. In: Advances in Neural Information Processing Systems, pp. 6965–6975 (2018)

2018
[25]

In: Advances in Neural Information Processing Systems, pp

Lakshminarayanan, B., Pritzel, A., Blundell, C.: Simple and scalable predictive uncertainty estimation using deep ensembles. In: Advances in Neural Information Processing Systems, pp. 6402–6413 (2017) 16

2017

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

Journal of Medical Systems 45(4), 1–16 (2021)

Alloqmani, Y., et al.: Deep learning approaches for detection and diagnosis of brain tumors: A survey. Journal of Medical Systems 45(4), 1–16 (2021)

work page 2021

[2] [2]

IEEE Access 9, 121156–121183 (2021)

Masood, A., et al.: A comprehensive survey of brain tumor detection in MRI: A critical review. IEEE Access 9, 121156–121183 (2021)

work page 2021

[3] [3]

Artificial Intelligence in Medicine 121, 102197 (2021)

Fernando, K.R.M., et al.: Deep learning for biomedical image reconstruction: A survey. Artificial Intelligence in Medicine 121, 102197 (2021)

work page 2021

[4] [4]

Medical Image Analysis 54, 10–19 (2019)

Zhang, J., et al.: Medical image classification using synergic deep learning. Medical Image Analysis 54, 10–19 (2019)

work page 2019

[5] [5]

ACM Computing Surveys 41(3), 1–58 (2009)

Chandola, V., Banerjee, A., Kumar, V.: Anomaly detection: A survey. ACM Computing Surveys 41(3), 1–58 (2009)

work page 2009

[6] [6]

Medical Image Analysis 8(3), 275–283 (2004)

Prastawa, M., Bullitt, E., Ho, S., Gerig, G.: A brain tumor segmentation framework based on outlier detection. Medical Image Analysis 8(3), 275–283 (2004)

work page 2004

[7] [7]

Special Lecture on IE 2(1), 1–18 (2015)

An, J., Cho, S.: Variational autoencoder based anomaly detection using recon- struction probability. Special Lecture on IE 2(1), 1–18 (2015)

work page 2015

[8] [8]

arXiv preprint arXiv:1812.05941 (2019)

Zimmerer, D., Kohl, S.A., Petersen, J., Isensee, F., Maier-Hein, K.H.: Context- encoding variational autoencoder for unsupervised anomaly detection. arXiv preprint arXiv:1812.05941 (2019)

arXiv 2019