Improving the Accuracy of Medical Diagnosis with Causal Machine Learning

Overview

Journal Nat Commun

Specialty Biology

Date 2020 Aug 13

PMID 32782264

Citations 86

Authors

Jonathan G Richens

Ciaran M Lee

Saurabh Johri

Affiliations

Soon will be listed here.

Abstract

Machine learning promises to revolutionize clinical decision making and diagnosis. In medical diagnosis a doctor aims to explain a patient's symptoms by determining the diseases causing them. However, existing machine learning approaches to diagnosis are purely associative, identifying diseases that are strongly correlated with a patients symptoms. We show that this inability to disentangle correlation from causation can result in sub-optimal or dangerous diagnoses. To overcome this, we reformulate diagnosis as a counterfactual inference task and derive counterfactual diagnostic algorithms. We compare our counterfactual algorithms to the standard associative algorithm and 44 doctors using a test set of clinical vignettes. While the associative algorithm achieves an accuracy placing in the top 48% of doctors in our cohort, our counterfactual algorithm places in the top 25% of doctors, achieving expert clinical accuracy. Our results show that causal reasoning is a vital missing ingredient for applying machine learning to medical diagnosis.

Citing Articles

Counterfactual explanations of tree based ensemble models for brain disease analysis with structure function coupling.

Wei S, Gao Z, Yao H, Qi X, Wang M, Huang J Sci Rep. 2025; 15(1):8524.

PMID: 40075142 PMC: 11904222. DOI: 10.1038/s41598-025-92316-x.

Diagnosis of Benign and Malignant Newly Developed Nodules on the Surgical Side After Breast Cancer Surgery Based on Machine Learning.

Wang Z, Li Q, Wang Y, Qian L, Hu X, Liu D Breast J. 2025; 2025:8511049.

PMID: 39996101 PMC: 11850066. DOI: 10.1155/tbj/8511049.

Causal machine learning models for predicting low birth weight in midwife-led continuity care intervention in North Shoa Zone, Ethiopia.

Moges W, Tegegne A, Mitku A, Tesfahun E, Hailemeskel S BMC Med Inform Decis Mak. 2025; 25(1):64.

PMID: 39920662 PMC: 11806756. DOI: 10.1186/s12911-025-02917-9.

Step-by-step causal analysis of EHRs to ground decision-making.

Doutreligne M, Struja T, Abecassis J, Morgand C, Celi L, Varoquaux G PLOS Digit Health. 2025; 4(2):e0000721.

PMID: 39899627 PMC: 11790099. DOI: 10.1371/journal.pdig.0000721.

MRI-based deep learning radiomics to differentiate dual-phenotype hepatocellular carcinoma from HCC and intrahepatic cholangiocarcinoma: a multicenter study.

Wu Q, Zhang T, Xu F, Cao L, Gu W, Zhu W Insights Imaging. 2025; 16(1):27.

PMID: 39881111 PMC: 11780023. DOI: 10.1186/s13244-025-01904-y.

References

Greenland S . For and Against Methodologies: Some Perspectives on Recent Causal and Statistical Inference Debates. Eur J Epidemiol. 2017; 32(1):3-20. DOI: 10.1007/s10654-017-0230-6. View

Badgeley M, Zech J, Oakden-Rayner L, Glicksberg B, Liu M, Gale W . Deep learning predicts hip fracture using confounding patient and healthcare variables. NPJ Digit Med. 2019; 2:31. PMC: 6550136. DOI: 10.1038/s41746-019-0105-1. View

Shwe M, Middleton B, Heckerman D, Henrion M, Horvitz E, Lehmann H . Probabilistic diagnosis using a reformulation of the INTERNIST-1/QMR knowledge base. I. The probabilistic model and inference algorithms. Methods Inf Med. 1991; 30(4):241-55. View

Graber M . The incidence of diagnostic error in medicine. BMJ Qual Saf. 2013; 22 Suppl 2:ii21-ii27. PMC: 3786666. DOI: 10.1136/bmjqs-2012-001615. View

Semigran H, Linder J, Gidengil C, Mehrotra A . Evaluation of symptom checkers for self diagnosis and triage: audit study. BMJ. 2015; 351:h3480. PMC: 4496786. DOI: 10.1136/bmj.h3480. View

Rizzi D . Causal reasoning and the diagnostic process. Theor Med. 1994; 15(3):315-33. DOI: 10.1007/BF01313345. View

Tomasev N, Glorot X, Rae J, Zielinski M, Askham H, Saraiva A . A clinically applicable approach to continuous prediction of future acute kidney injury. Nature. 2019; 572(7767):116-119. PMC: 6722431. DOI: 10.1038/s41586-019-1390-1. View

Cournoyea M, Kennedy A . Causal explanatory pluralism and medically unexplained physical symptoms. J Eval Clin Pract. 2014; 20(6):928-33. DOI: 10.1111/jep.12238. View

Miller R . A history of the INTERNIST-1 and Quick Medical Reference (QMR) computer-assisted diagnosis projects, with lessons learned. Yearb Med Inform. 2010; :121-36. View

10.

Litjens G, Sanchez C, Timofeeva N, Hermsen M, Nagtegaal I, Kovacs I . Deep learning as a tool for increased accuracy and efficiency of histopathological diagnosis. Sci Rep. 2016; 6:26286. PMC: 4876324. DOI: 10.1038/srep26286. View

11.

Heckerman D, Horvitz E, Nathwani B . Toward normative expert systems: Part I. The Pathfinder project. Methods Inf Med. 1992; 31(2):90-105. View

12.

Greenland S, Pearl J, Robins J . Causal diagrams for epidemiologic research. Epidemiology. 1999; 10(1):37-48. View

13.

Jiang F, Jiang Y, Zhi H, Dong Y, Li H, Ma S . Artificial intelligence in healthcare: past, present and future. Stroke Vasc Neurol. 2018; 2(4):230-243. PMC: 5829945. DOI: 10.1136/svn-2017-000101. View

14.

Yu K, Beam A, Kohane I . Artificial intelligence in healthcare. Nat Biomed Eng. 2019; 2(10):719-731. DOI: 10.1038/s41551-018-0305-z. View

15.

Heckerman D, Nathwani B . Toward normative expert systems: Part II. Probability-based representations for efficient knowledge acquisition and inference. Methods Inf Med. 1992; 31(2):106-16. View

16.

Silver D, Schrittwieser J, Simonyan K, Antonoglou I, Huang A, Guez A . Mastering the game of Go without human knowledge. Nature. 2017; 550(7676):354-359. DOI: 10.1038/nature24270. View

17.

Cooper G, Abraham V, Aliferis C, Aronis J, Buchanan B, Caruana R . Predicting dire outcomes of patients with community acquired pneumonia. J Biomed Inform. 2005; 38(5):347-66. DOI: 10.1016/j.jbi.2005.02.005. View

18.

Liberman A, Newman-Toker D . Symptom-Disease Pair Analysis of Diagnostic Error (SPADE): a conceptual framework and methodological approach for unearthing misdiagnosis-related harms using big data. BMJ Qual Saf. 2018; 27(7):557-566. PMC: 6049698. DOI: 10.1136/bmjqs-2017-007032. View

19.

Converse L, Barrett K, Rich E, Reschovsky J . Methods of Observing Variations in Physicians' Decisions: The Opportunities of Clinical Vignettes. J Gen Intern Med. 2015; 30 Suppl 3:S586-94. PMC: 4512963. DOI: 10.1007/s11606-015-3365-8. View

20.

Dresselhaus T, Peabody J, Luck J, Bertenthal D . An evaluation of vignettes for predicting variation in the quality of preventive care. J Gen Intern Med. 2004; 19(10):1013-8. PMC: 1492573. DOI: 10.1007/s11606-004-0003-2. View