Rao, S., Li, Y., Ramakrishnan, R., Hassaine, A., Canoy, D., Cleland, J. G. , Lukasiewicz, T., Salimi-Khorshidi, G. and Rahimi, K. (2022) An explainable transformer-based deep learning model for the prediction of incident heart failure. IEEE Journal of Biomedical and Health Informatics, 26(7), pp. 3362-3372. (doi: 10.1109/JBHI.2022.3148820) (PMID:35130176)
![[img]](https://eprints.gla.ac.uk/style/images/fileicons/text.png) |
Text
265449.pdf - Published Version Available under License Creative Commons Attribution. 5MB |
Abstract
Predicting the incidence of complex chronic conditions such as heart failure is challenging. Deep learning models applied to rich electronic health records may improve prediction but remain unexplainable hampering their wider use in medical practice. We aimed to develop a deep-learning framework for accurate and yet explainable prediction of 6-month incident heart failure (HF). Using 100,071 patients from longitudinal linked electronic health records across the UK, we applied a novel Transformer-based risk model using all community and hospital diagnoses and medications contextualized within the age and calendar year for each patient's clinical encounter. Feature importance was investigated with an ablation analysis to compare model performance when alternatively removing features and by comparing the variability of temporal representations. A post-hoc perturbation technique was conducted to propagate the changes in the input to the outcome for feature contribution analyses. Our model achieved 0.93 area under the receiver operator curve and 0.69 area under the precision-recall curve on internal 5-fold cross validation and outperformed existing deep learning models. Ablation analysis indicated medication is important for predicting HF risk, calendar year is more important than chronological age, which was further reinforced by temporal variability analysis. Contribution analyses identified risk factors that are closely related to HF. Many of them were consistent with existing knowledge from clinical and epidemiological research but several new associations were revealed which had not been considered in expert-driven risk prediction models. In conclusion, the results highlight that our deep learning model, in addition high predictive performance, can inform data-driven risk factor identification.
| Item Type: | Articles |
|---|---|
| Additional Information: | The following authors are supported by grants from the British Heart Foundation (BHF): YL & KR (grant number: FS/PhD/21/29110) and KR & DC (grant number: PG/18/65/33872); KR is also in receipt of funding from the UKRI’s Global Challenges Research Fund (GCRF), Grant Ref ES/P0110551/1, Oxford National Institute of Health Research (NIHR) Biomedical Research Centre and the Oxford Martin School (OMS), University of Oxford. JGC is in receipt of funding from NIHR (NIHRDH-NIHR130487) and BHF (RE/18/6/34217). TL is in receipt of funding from the Alan Turing Institute (ATI) under the EPSRC grant (EP/N510129/1), by the AXA Research Fund, and by the EU TAILOR grant. |
| Status: | Published |
| Refereed: | Yes |
| Glasgow Author(s) Enlighten ID: | Cleland, Professor John |
| Authors: | Rao, S., Li, Y., Ramakrishnan, R., Hassaine, A., Canoy, D., Cleland, J. G., Lukasiewicz, T., Salimi-Khorshidi, G., and Rahimi, K. |
| Subjects: | R Medicine > R Medicine (General) |
| College/School: | College of Medical Veterinary and Life Sciences > School of Health & Wellbeing > Robertson Centre |
| Journal Name: | IEEE Journal of Biomedical and Health Informatics |
| Publisher: | IEEE |
| ISSN: | 2168-2194 |
| ISSN (Online): | 2168-2208 |
| Published Online: | 07 February 2022 |
| Copyright Holders: | Copyright © 2022 The Authors |
| First Published: | First published in IEEE Journal of Biomedical and Health Informatics 26(7): 3362-3372 |
| Publisher Policy: | Reproduced under a Creative Commons Licence |
University Staff: Request a correction | Enlighten Editors: Update this record
Funder and Project Information
Funder and Project Information