Causality matters in medical imaging

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.

Full text

Its licence does not allow us to show the full text. Read it at PubMed Central:

References 67

1. 1.

   Esteva A, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;542:115-118.

   Not in our library10.1038/nature21056

2. 2.

   Menze BH, et al. The multimodal brain tumor image segmentation benchmark (BRATS). IEEE Trans. Med. Imag. 2015;34:1993-2024.

   Not in our library10.1109/tmi.2014.2377694

3. 3.

   Bareinboim E, Pearl J. Causal inference and the data-fusion problem. Proc. Natl Acad. Sci. USA. 2016;113:7345-7352.

   Not in our library10.1073/pnas.1510507113

4. 4.

   Lucas PJF, van der Gaag LC, Abu-Hanna A. Bayesian networks in biomedicine and health-care. Artif. Intell. Med. 2004;30:201-214.

   Not in our library10.1016/j.artmed.2003.11.001

5. 5.

   Cypko MA, et al. Validation workflow for a clinical Bayesian network model in multidisciplinary decision making in head and neck oncology treatment. Int. J. Computer Assist. Radiol. Surg. 2017;12:1959-1970.

   Not in our library10.1007/s11548-017-1531-7

6. 6.

   Dwork C, et al. The reusable holdout: preserving validity in adaptive data analysis. Science. 2015;349:636-638.

   Not in our library10.1126/science.aaa9375

7. 7.

   Schölkopf, B. et al. On causal and anticausal learning. In Proc. 29th International Conference on Machine Learning (ICML 2012) 459–466 (2012).

   Not in our library

8. 8.

   Chapelle, O., Schölkopf, B. & Zien, A. (eds) Semi-Supervised Learning (MIT Press, Cambridge, MA, 2006).

   Not in our library

9. 9.

   Schölkopf, B., et al. in Empirical Inference (eds Schölkopf, B., Luo, Z. & Vovk, V.) Ch. 13, 129–141 (Springer, Berlin, Heidelberg, 2013).

   Not in our library

10. 10.

    Blöbaum, P., Shimizu, S. & Washio, T. in Advanced Methodologies for Bayesian Networks (AMBN 2015) Vol. 9505, 209–221 (Springer, Cham, 2015).

    Not in our library

11. 11.
    Why rankings of biomedical image analysis competitions should be interpreted with care · 2018

    Corrected
12. 12.

    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).

    Not in our library

13. 13.

    Cozman, F. & Cohen, I. in Semisupervised Learning (eds Chapelle, O. et al.) Ch. 4, 57–72 (MIT Press, Cambridge, MA, 2006).

    Not in our library

14. 14.

    Singh, A., Nowak, R. & Zhu, X. in Advances in Neural Information Processing Systems Vol. 21 (NIPS 2008), 1513–1520 (2008).

    Not in our library

15. 15.

    Balcan, M.-F. & Blum, A. in Semisupervised Learning (eds Chapelle, O. et al.) Ch. 22, 397–419 (MIT Press, Cambridge, MA, 2006).

    Not in our library

16. 16.

    Quiñonero-Candela, J., Sugiyama, M., Schwaighofer, A. & Lawrence, N. D. (eds.) Dataset Shift in Machine Learning (MIT Press, Cambridge, MA, 2009).

    Not in our library

17. 17.

    Pearl J, Bareinboim E. External validity: from do-calculus to transportability across populations. Stat. Sci. 2014;29:579-595.

    Not in our library10.1214/14-sts486

18. 18.

    Storkey, A. J. in Dataset Shift in Machine Learning (eds Quiñonero-Candela, J. et al.), Ch. 1, 3–28 (MIT Press, Cambridge, MA, 2009).

    Not in our library

19. 19.

    Zhang, K., Gong, M. & Schölkopf, B. Multi-source domain adaptation: a causal view. In Proc. Twenty-Ninth AAAI Conference on Artificial Intelligence (AAAI 2015) 3150–3157 (AAAI, 2015).

    Not in our library

20. 20.

    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).

    Not in our library

21. 21.

    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).

    Not in our library

22. 22.

    Moreno-Torres JG, Raeder T, Alaiz-Rodríguez R, Chawla NV, Herrera F. A unifying view on dataset shift in classification. Pattern Recognit. 2012;45:521-530.

    Not in our library10.1016/j.patcog.2011.06.019

23. 23.

    Glocker, B., Robinson, R., Castro, D. C., Dou, Q. & Konukoglu, E. Machine learning with multi-site imaging data: an empirical study on the impact of scanner effects. Preprint at https://arxiv.org/abs/1910.04597 (2019).

    Not in our library

24. 24.

    Ganin Y, et al. Domain-adversarial training of neural networks. J. Mach. Learn. Res. 2016;17:1-35.

    Not in our library

25. 25.

    Kamnitsas, K. et al. in Information Processing in Medical Imaging, Vol. 10265, 597–609 (Springer, Cham, 2017).

    Not in our library

26. 26.

    Frangi AF, Tsaftaris SA, Prince JL. Simulation and synthesis in medical imaging. IEEE Trans. Med. Imag. 2018;37:673-679.

    Not in our library10.1109/tmi.2018.2800298

27. 27.

    Huo Y, et al. SynSeg-Net: synthetic segmentation without target modality ground truth. IEEE Trans. Med. Imag. 2019;38:1016-1025.

    Not in our library10.1109/tmi.2018.2876633

28. 28.

    Arjovsky, M., Bottou, L., Gulrajani, I. & Lopez-Paz, D. Invariant risk minimization. Preprint at https://arxiv.org/abs/1907.02893 (2019).

    Not in our library

29. 29.

    Bareinboim, E., Tian, J. & Pearl, J., Recovering from selection bias in causal and statistical inference. In Proc. Twenty-Eighth AAAI Conference on Artificial Intelligence (AAAI 2014) 2410–2416 (AAAI, 2014).

    Not in our library

30. 30.

    Hernán MA, Hernández-Díaz S, Robins JM. A structural approach to selection bias. Epidemiology. 2004;15:615-625.

    Not in our library10.1097/01.ede.0000135174.63482.43

31. 31.

    Pearl, J. Causality: Models, Reasoning, and Inference 2nd edn (Cambridge University Press, Cambridge, UK, 2009).

    Not in our library

32. 32.

    Hernán MA, Cole SR. Causal diagrams and measurement bias. Am. J. Epidemiol. 2009;170:959-962.

    Not in our library10.1093/aje/kwp293

33. 33.

    Shahar E. Causal diagrams for encoding and evaluation of information bias. J. Eval. Clin. Pract. 2009;15:436-440.

    Not in our library10.1111/j.1365-2753.2008.01031.x

34. 34.

    Lash TL, et al. Good practices for quantitative bias analysis. Int. J. Epidemiol. 2014;43:1969-1985.

    Not in our library10.1093/ije/dyu149

35. 35.

    Rubin DB. Inference and missing data. Biometrika. 1976;63:581.

    Not in our library10.1093/biomet/63.3.581

36. 36.

    Daniel RM, Kenward MG, Cousens SN, De Stavola BL. Using causal diagrams to guide analysis in missing data problems. Stat. Methods Med. Res. 2012;21:243-256.

    Not in our library10.1177/0962280210394469

37. 37.

    Mohan, K., Pearl, J. & Tian, J. in Advances in Neural Information Processing Systems Vol. 26 (NIPS 2013), 1277–1285 (2013).

    Not in our library

38. 38.

    Miller KL, et al. Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nat. Neurosci. 2016;19:1523-1536.

    Not in our library10.1038/nn.4393

39. 39.

    Conroy M, et al. The advantages of UK Biobank’s open-access strategy for health research. J. Intern. Med. 2019;286:389-397.

    Not in our library10.1111/joim.12955

40. 40.

    US Food and Drug Administration. Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD). https://www.fda.gov/media/122535/download (FDA, 2019).

    Not in our library

Cited by papers we hold 3

• Artificial intelligence and machine learning in cancer imaging · 2022
  No notice
• Surgical data science – from concepts toward clinical translation · 2022
  No notice
• Clinical artificial intelligence quality improvement: towards continual monitoring and updating of AI algorithms in healthcare · 2022
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.