CoChat

Causality matters in medical imaging

Open access
Castro, Daniel C., Walker, Ian, Glocker, Ben · 2020 · Nature Communications

The DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index.

Abstract

Causal reasoning can shed new light on the major challenges in machine learning for medical imaging: scarcity of high-quality annotated data and mismatch between the development dataset and the target environment. A causal perspective on these issues allows decisions about data collection, annotation, preprocessing, and learning strategies to be made and scrutinized more transparently, while providing a detailed categorisation of potential biases and mitigation techniques. Along with worked clinical examples, we highlight the importance of establishing the causal relationship between images and their annotations, and offer step-by-step recommendations for future studies.

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 · 10.1038/s41467-020-17478-w

Cites

54 of 54 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Key challenges in machine learning for medical imaging. a Data scarcity and b–d data mismatch. X represents images and Y, annotations (e.g. diagnosis labels). Ptr refers to the distribution of data available for training a predictive model, and Pte is the test distribution, i.e. data that will be encountered once the model is deployed. Dots represent data points with any label, while circles and crosses indicate images with different labels (e.g. cases vs. controls).
Fig. 1 Key challenges in machine learning for medical imaging. a Data scarcity and b–d data mismatch. X represents images and Y, annotations (e.g. diagnosis labels). Ptr refers to the distribution of data available for training a predictive model, and Pte is the test distribution, i.e. data that will be encountered once the model is deployed. Dots represent data points with any label, while circles and crosses indicate images with different labels (e.g. cases vs. controls).CC BY · PubMed Central
Fig. 2 Causal diagrams for medical imaging examples. a Skin lesion classification. b Prostate tumour segmentation. Filled circular nodes represent measured variables, double circular nodes denote sample selection indicators, and squares are used for sample domain indicators. Here we additionally highlight the direction of the predictive task.
Fig. 2 Causal diagrams for medical imaging examples. a Skin lesion classification. b Prostate tumour segmentation. Filled circular nodes represent measured variables, double circular nodes denote sample selection indicators, and squares are used for sample domain indicators. Here we additionally highlight the direction of the predictive task.CC BY · PubMed Central
Fig. 3 Selection diagrams for dataset shift. a–c Causal and d–f anticausal scenarios, with corresponding factorisations of the joint distribution PD(X, Y, Z). X is the acquired image; Y, the prediction target; Z, the unobserved true anatomy; and D, the domain indicator (0: ‘train’, 1: ‘test’). An unfilled node means the variable is unmeasured.
Fig. 3 Selection diagrams for dataset shift. a–c Causal and d–f anticausal scenarios, with corresponding factorisations of the joint distribution PD(X, Y, Z). X is the acquired image; Y, the prediction target; Z, the unobserved true anatomy; and D, the domain indicator (0: ‘train’, 1: ‘test’). An unfilled node means the variable is unmeasured.CC BY · PubMed Central
Fig. 4 Causal diagrams for different sample selection scenarios. a Random; b image-dependent; c target-dependent; d jointly dependent. S = 1 indicates an observed sample, and plain edges represent either direction.
Fig. 4 Causal diagrams for different sample selection scenarios. a Random; b image-dependent; c target-dependent; d jointly dependent. S = 1 indicates an observed sample, and plain edges represent either direction.CC BY · PubMed Central
Fig. 5 A ‘scaffold’ causal diagram summarising typical medical imaging workflows. We believe most practical cases can be adapted from this generic structure by removing or adding elements. Here are represented a variety of possible prediction targets (marked Y1–Y4): some anticausal (Y1, Y2) and others, causal (Y3, Y4). ‘Annotation’ here refers to any image-derived data, such as lesion descriptions, regions of interest, spatial landmark coordinates, or segmentation maps. Note that annotators will often be aware of the patients' records and diagnoses, in which cases there could be additional arrows from Y1 or Y2 towards Y4.
Fig. 5 A ‘scaffold’ causal diagram summarising typical medical imaging workflows. We believe most practical cases can be adapted from this generic structure by removing or adding elements. Here are represented a variety of possible prediction targets (marked Y1–Y4): some anticausal (Y1, Y2) and others, causal (Y3, Y4). ‘Annotation’ here refers to any image-derived data, such as lesion descriptions, regions of interest, spatial landmark coordinates, or segmentation maps. Note that annotators will often be aware of the patients' records and diagnoses, in which cases there could be additional arrows from Y1 or Y2 towards Y4.CC BY · PubMed Central

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

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

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 54 references

Research works printed but not linked: 9

  1. Schölkopf, B. et al. On causal and anticausal learning. In Proc. 29th International Conference on Machine Learning (ICML 2012) 459–466 (2012).
  2. Oliver, A., Odena, A., Raffel, C. A., Cubuk, E. D. & Goodfellow, I. in Advances in Neural Information Processing Systems Vol. 31 (NeurIPS 2018), 3235–3246 (2018).
  3. Singh, A., Nowak, R. & Zhu, X. in Advances in Neural Information Processing Systems Vol. 21 (NIPS 2008), 1513–1520 (2008).
  4. Storkey, A. J. in Dataset Shift in Machine Learning (eds Quiñonero-Candela, J. et al.), Ch. 1, 3–28 (MIT Press, Cambridge, MA, 2009).
  5. Zhang, K., Schölkopf, B., Muandet, K. & Wang, Z. Domain adaptation under target and conditional shift. In Proc. 30th International Conference on Machine Learning (ICML 2013), Vol. 28, 819–827 (PMLR, 2013).
  6. Lipton, Z. C., Wang, Y.-X. & Smola, A. J. Detecting and correcting for label shift with black box predictors. In Proc. 35th International Conference on Machine Learning (ICML 2018), Vol. 80, 3122–3130 (PMLR, 2018).
  7. Mohan, K., Pearl, J. & Tian, J. in Advances in Neural Information Processing Systems Vol. 26 (NIPS 2013), 1277–1285 (2013).
  8. Subbaswamy, A., Schulam, P. & Saria, S. Preventing failures due to dataset shift: Learning predictive models that transport. In Proc. Twenty-Second International Conference on Artificial Intelligence and Statistics (AISTATS 2019), Vol. 89, 3118–3127 (PMLR, 2019).
  9. Peters, J., Janzing, D. & Schölkopf, B. Elements of Causal Inference: Foundations and Learning Algorithms (MIT Press, Cambridge, MA, 2017).

Cited by

See all 452

Metadata

Authors
Castro, D. C., Walker, I., & Glocker, B.
Affiliations
Imperial College London (GB)
Corresponding author
  • Castro, Daniel C.
  • Glocker, Ben
Venue
Nature Communications
Published
2020-07-22
Volume
11
Issue
1
Pages
3673-3673
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2995750307
Cited by
364
References
49
Citations by year
2026: 70 · 2025: 77 · 2024: 59 · 2023: 72 · 2022: 52 · 2021: 33 · 2020: 1
FWCI
3.02
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Spiral (Imperial College London) (submittedVersion, other-oa)
  • Spiral (Imperial College London) (submittedVersion, cc-by)
  • arXiv (Cornell University)
Funders
  • European Commission (ERC-2017, 757173, HORIZON2020)
  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
  • Natural Environment Research Council (NE/L002515/1)
Topics
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
  • Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
Keywords
causal reasoning, diagnostic imaging, annotated data scarcity, domain shift, annotation bias, image annotation, bias mitigation, data collection, image preprocessing, learning strategies
MeSH terms
Machine Learning, Diagnostic Imaging, Humans, Image Interpretation, Computer-Assisted, Causality

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: Causality matters in medical imaging. https://app.cochat.ai/papers/10.1038/s41467-020-17478-w
The paper
Castro, D. C., Walker, I., & Glocker, B. (2020). Causality matters in medical imaging. Nature Communications. https://doi.org/10.1038/s41467-020-17478-w

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.