CoChat

Explanation and Use of Uncertainty Quantified by Bayesian Neural Network Classifiers for Breast Histopathology Images

Open access
Thiagarajan, Ponkrshnan, Khairnar, Pushkar, Ghosh, Susanta · 2022 · IEEE Transactions on Medical Imaging

Published version of a preprint

Cited by 57

The DOI matches its Crossref record; 14 of the 14 references it cites were found in a registry or an index.

Abstract

Despite the promise of Convolutional neural network (CNN) based classification models for histopathological images, it is infeasible to quantify its uncertainties. Moreover, CNNs may suffer from overfitting when the data is biased. We show that Bayesian-CNN can overcome these limitations by regularizing automatically and by quantifying the uncertainty. We have developed a novel technique to utilize the uncertainties provided by the Bayesian-CNN that significantly improves the performance on a large fraction of the test data (about 6% improvement in accuracy on 77% of test data). Further, we provide a novel explanation for the uncertainty by projecting the data into a low dimensional space through a nonlinear dimensionality reduction technique. This dimensionality reduction enables interpretation of the test data through visualization and reveals the structure of the data in a low dimensional feature space. We show that the Bayesian-CNN can perform much better than the state-of-the-art transfer learning CNN (TL-CNN) by reducing the false negative and false positive by 11% and 7.7% respectively for the present data set. It achieves this performance with only 1.86 million parameters as compared to 134.33 million for TL-CNN. Besides, we modify the Bayesian-CNN by introducing a stochastic adaptive activation function. The modified Bayesian-CNN performs slightly better than Bayesian-CNN on all performance metrics and significantly reduces the number of false negatives and false positives (3% reduction for both). We also show that these results are statistically significant by performing McNemar's statistical significance test. This work shows the advantages of Bayesian-CNN against the state-of-the-art, explains and utilizes the uncertainties for histopathological images. It should find applications in various medical image classifications.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/tmi.2021.3123300

Cites

14 of 14 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 14 references

Research works printed but not linked: 55

  1. Wild, 2014, World Cancer Report 2014
  2. 10.1001/jamaoncol.2018.2706 · DOI · Paper page
  3. 10.1186/bcr932 · DOI · Paper page
  4. 10.1038/s41598-017-04075-z · DOI · Paper page
  5. 10.1146/annurev-pathol-011811-120902 · DOI · Paper page
  6. 10.1111/j.1365-2559.2011.03814.x · DOI · Paper page
  7. 10.1016/j.compmedimag.2011.02.006 · DOI · Paper page
  8. 10.1016/j.imu.2016.11.001 · DOI · Paper page
  9. 10.1109/TMI.2013.2275151 · DOI · Paper page
  10. 10.1109/JSYST.2013.2279415 · DOI · Paper page
  11. 10.1016/j.compbiomed.2013.08.003 · DOI · Paper page
  12. 10.1016/j.procs.2016.04.224 · DOI · Paper page
  13. 10.1007/s00138-012-0459-8 · DOI
  14. 10.1109/TBME.2015.2496264 · DOI · Paper page
  15. 10.1109/TMI.2017.2758580 · DOI · Paper page
  16. 10.1109/ICPR.2016.7900002 · DOI · Paper page
  17. 10.1117/12.2043872 · DOI · Paper page
  18. 10.1109/IJCNN.2016.7727519 · DOI · Paper page
  19. 10.1371/journal.pone.0177544 · DOI · Paper page
  20. 10.1109/TMI.2019.2891305 · DOI · Paper page
  21. 10.1109/TPAMI.2016.2644615 · DOI · Paper page
  22. 10.1109/ACCESS.2019.2898044 · DOI · Paper page
  23. 10.1109/TMI.2019.2948026 · DOI · Paper page
  24. 10.1109/TMI.2019.2962013 · DOI
  25. 10.14569/IJACSA.2018.090645 · DOI
  26. Motlagh, 2018, Breast cancer histopathological image classification: A deep learning approach, Biorxiv
  27. 10.3389/fgene.2019.00080 · DOI · Paper page
  28. 10.1016/j.media.2019.05.010 · DOI · Paper page
  29. 10.1007/978-3-319-93000-8_91 · DOI
  30. 10.1007/978-3-319-93000-8_106 · DOI
  31. 10.1007/978-3-319-93000-8_87 · DOI
  32. 10.1109/CVPR.2017.243 · DOI · Paper page
  33. 10.1109/CVPR.2015.7298594 · DOI · Paper page
  34. Simonyan, 2014, Very deep convolutional networks for large-scale image recognition, arXiv:1409.1556 · arXiv
  35. 10.5555/2999134.2999257 · doi:10.5555/2999134.2999257
  36. Shridhar, 2019, A comprehensive guide to Bayesian convolutional neural network with variational inference, arXiv:1901.02731 · arXiv
  37. Shridhar, 2018, Uncertainty estimations by softplus normalization in Bayesian convolutional neural networks with variational inference, arXiv:1806.05978 · arXiv
  38. 10.5555/3295222.3295309 · doi:10.5555/3295222.3295309
  39. Kwon, Uncertainty quantification using Bayesian neural networks in classification: Application to ischemic stroke lesion segmentation
  40. 10.1038/s41598-019-50587-1 · DOI · Paper page
  41. 10.1016/j.neucom.2017.06.070 · DOI · Paper page
  42. 10.1016/j.jcp.2019.109136 · DOI · Paper page
  43. 10.2352/ISSN.2470-1173.2016.19.COIMG-149 · DOI
  44. Mooney, 2017, Breast Histopathology Images
  45. Van der Maaten, 2008, Visualizing data using t-SNE, J. Mach. Learn. Res., 9, 1
  46. Bergstra, Algorithms for hyper-parameter optimization, 24, 1
  47. Bergstra, Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures, 115
  48. 10.1109/TMI.2016.2528162 · DOI · Paper page
  49. Bishop, 2006, Pattern Recognition and Machine Learning
  50. Orr, 2003, Neural Networks: Tricks of the Trade
  51. Friedman, 2001, The Elements of Statistical Learning, 1
  52. LeCun, 1998, The MNIST Database of Handwritten Digits
  53. 10.1177/001316446002000104 · DOI · Paper page
  54. 10.1371/journal.pone.0202121 · DOI · Paper page
  55. 10.1371/journal.pone.0242301 · DOI

Cited by

See all 57

Metadata

Authors
Thiagarajan, P., Khairnar, P., & Ghosh, S.
Venue
IEEE Transactions on Medical Imaging
Published
2022-04
Volume
41
Issue
4
Pages
815-825
ISSN
  • 0278-0062 (ISSN-L)
  • 1558-254X
OpenAlex
W3211109300
Cited by
57
References
14
FWCI
2.82
Type
journal article
Language
en
Open access
gold
Funders
  • High Performance Computing (HPC) Facilities through SUPERIOR at Michigan Technological University
  • Extreme Science and Engineering Discovery Environment (MSS200004)

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Explanation and Use of Uncertainty Quantified by Bayesian Neural Network Classifiers for Breast Histopathology Images. https://app.cochat.ai/papers/10.1109/tmi.2021.3123300
The paper
Thiagarajan, P., Khairnar, P., & Ghosh, S. (2022). Explanation and Use of Uncertainty Quantified by Bayesian Neural Network Classifiers for Breast Histopathology Images. IEEE Transactions on Medical Imaging. https://doi.org/10.1109/tmi.2021.3123300

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.