Papers
Topics
Authors
Recent
Search
2000 character limit reached

Assisting Clinical Decisions for Scarcely Available Treatment via Disentangled Latent Representation

Published 6 Jul 2023 in cs.LG and cs.AI | (2307.03315v1)

Abstract: Extracorporeal membrane oxygenation (ECMO) is an essential life-supporting modality for COVID-19 patients who are refractory to conventional therapies. However, the proper treatment decision has been the subject of significant debate and it remains controversial about who benefits from this scarcely available and technically complex treatment option. To support clinical decisions, it is a critical need to predict the treatment need and the potential treatment and no-treatment responses. Targeting this clinical challenge, we propose Treatment Variational AutoEncoder (TVAE), a novel approach for individualized treatment analysis. TVAE is specifically designed to address the modeling challenges like ECMO with strong treatment selection bias and scarce treatment cases. TVAE conceptualizes the treatment decision as a multi-scale problem. We model a patient's potential treatment assignment and the factual and counterfactual outcomes as part of their intrinsic characteristics that can be represented by a deep latent variable model. The factual and counterfactual prediction errors are alleviated via a reconstruction regularization scheme together with semi-supervision, and the selection bias and the scarcity of treatment cases are mitigated by the disentangled and distribution-matched latent space and the label-balancing generative strategy. We evaluate TVAE on two real-world COVID-19 datasets: an international dataset collected from 1651 hospitals across 63 countries, and a institutional dataset collected from 15 hospitals. The results show that TVAE outperforms state-of-the-art treatment effect models in predicting both the propensity scores and factual outcomes on heterogeneous COVID-19 datasets. Additional experiments also show TVAE outperforms the best existing models in individual treatment effect estimation on the synthesized IHDP benchmark dataset.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (50)
  1. Integrating machine learning predictions for perioperative risk management: Towards an empirical design of a flexible-standardized risk assessment tool. Journal of Biomedical Informatics 137 (2023), 104270.
  2. ECMO during the COVID-19 pandemic: when is it unjustified? , 3 pages.
  3. Deep counterfactual networks with propensity-dropout. arXiv preprint arXiv:1706.05966 (2017).
  4. ECMO in COVID-19 patients: a systematic review and meta-analysis. Journal of cardiothoracic and vascular anesthesia (2021).
  5. Isolating sources of disentanglement in variational autoencoders. arXiv preprint arXiv:1802.04942 (2018).
  6. BART: Bayesian additive regression trees. The Annals of Applied Statistics 4, 1 (2010), 266–298.
  7. Really doing great at estimating CATE? a critical look at ML benchmarking practices in treatment effect estimation. In Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 2).
  8. Alicia Curth and Mihaela van der Schaar. 2021a. Nonparametric estimation of heterogeneous treatment effects: From theory to learning algorithms. In International Conference on Artificial Intelligence and Statistics. PMLR, 1810–1818.
  9. Alicia Curth and Mihaela van der Schaar. 2021b. On inductive biases for heterogeneous treatment effect estimation. Advances in Neural Information Processing Systems 34 (2021), 15883–15894.
  10. Extracorporeal membrane oxygenation for critically ill patients with COVID-19–related acute respiratory distress syndrome: worth the effort? American journal of respiratory and critical care medicine 202, 3 (2020), 460–463.
  11. Association Between Availability of ECMO and Mortality in COVID-19 Patients Eligible for ECMO: A Natural Experiment. American Journal of Respiratory and Critical Care Medicine ja (2022).
  12. Why do tree-based models still outperform deep learning on tabular data? arXiv preprint arXiv:2207.08815 (2022).
  13. Bayesian regression tree models for causal inference: Regularization, confounding, and heterogeneous effects (with discussion). Bayesian Analysis 15, 3 (2020), 965–1056.
  14. Role of extracorporeal membrane oxygenation in COVID-19: a systematic review. Journal of Cardiac Surgery 35, 10 (2020), 2679–2687.
  15. beta-vae: Learning basic visual concepts with a constrained variational framework. In International conference on learning representations.
  16. The role of extracorporeal membrane oxygenation in critically ill patients with COVID-19: a narrative review. BMC pulmonary medicine 21, 1 (2021), 1–13.
  17. Continuous real-time prediction of surgical case duration using a modular artificial neural network. British journal of anaesthesia 128, 5 (2022), 829–837.
  18. Learning representations for counterfactual inference. In International conference on machine learning. PMLR, 3020–3029.
  19. Self-explaining Hierarchical Model for Intraoperative Time Series. arXiv preprint arXiv:2210.04417 (2022).
  20. Predicting Intraoperative Hypoxemia with Hybrid Inference Sequence Autoencoder Networks. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management. 1269–1278.
  21. Causal effect inference with deep latent-variable models. Advances in neural information processing systems 30 (2017).
  22. ELSO Org. 2022. ELSO COVID-19 Dashboard. https://www.elso.org/COVID19.aspx
  23. World Health Organization et al. 2021. COVID-19 clinical management: living guidance, 25 January 2021. Technical Report. World Health Organization.
  24. Predicting mortality risk in patients undergoing venovenous ECMO for ARDS due to influenza A (H1N1) pneumonia: the ECMOnet score. Intensive care medicine 39, 2 (2013), 275–281.
  25. Paul R Rosenbaum and Donald B Rubin. 1983. The central role of the propensity score in observational studies for causal effects. Biometrika 70, 1 (1983), 41–55.
  26. The first and second waves of the COVID-19 pandemic in Africa: a cross-sectional study. The Lancet 397, 10281 (2021), 1265–1275.
  27. Data augmentation using generative adversarial networks (CycleGAN) to improve generalizability in CT segmentation tasks. Scientific reports 9, 1 (2019), 1–9.
  28. Predicting survival after extracorporeal membrane oxygenation for severe acute respiratory failure. The Respiratory Extracorporeal Membrane Oxygenation Survival Prediction (RESP) score. American journal of respiratory and critical care medicine 189, 11 (2014), 1374–1382.
  29. The PRESERVE mortality risk score and analysis of long-term outcomes after extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. Intensive care medicine 39, 10 (2013), 1704–1713.
  30. Neel Shah and Ahmed S Said. 2021. Extracorporeal Support Prognostication—Time to Move the Goal Posts? Membranes 11, 7 (2021), 537.
  31. Validation of extracorporeal membrane oxygenation mortality prediction & severity of illness scores in an international COVID-19 cohort. Artificial Organs (2023).
  32. Estimating individual treatment effect: generalization bounds and algorithms. In International Conference on Machine Learning. PMLR, 3076–3085.
  33. Extracorporeal life support organization coronavirus disease 2019 interim guidelines: a consensus document from an international group of interdisciplinary extracorporeal membrane oxygenation providers. Asaio Journal (2020).
  34. Adapting neural networks for the estimation of treatment effects. Advances in neural information processing systems 32 (2019).
  35. Obstructive Sleep Apnea Detection Using Difference in Feature and Modified Minimum Distance Classifier. In 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE, 1–4.
  36. Extracorporeal membrane oxygenation for coronavirus disease 2019-related acute respiratory distress syndrome. Current opinion in critical care 28, 1 (2022), 90.
  37. Learning Structured Output Representation using Deep Conditional Generative Models. In Advances in Neural Information Processing Systems, C. Cortes, N. Lawrence, D. Lee, M. Sugiyama, and R. Garnett (Eds.), Vol. 28. Curran Associates, Inc. https://proceedings.neurips.cc/paper/2015/file/8d55a249e6baa5c06772297520da2051-Paper.pdf
  38. Peter Spirtes. 2009. A Tutorial On Causal Inference. (2009).
  39. Wasserstein auto-encoders. arXiv preprint arXiv:1711.01558 (2017).
  40. Minimax estimation of maximum mean discrepancy with radial kernels. Advances in Neural Information Processing Systems 29 (2016).
  41. Stefan Wager and Susan Athey. 2018. Estimation and inference of heterogeneous treatment effects using random forests. J. Amer. Statist. Assoc. 113, 523 (2018), 1228–1242.
  42. Deep generative model for robust imbalance classification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 14124–14133.
  43. Pengzhou Wu and Kenji Fukumizu. 2021. Intact-VAE: Estimating treatment effects under unobserved confounding. arXiv preprint arXiv:2101.06662 (2021).
  44. Perioperative Predictions with Interpretable Latent Representation. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery & Data Mining.
  45. Use of machine learning to develop and evaluate models using preoperative and intraoperative data to identify risks of postoperative complications. JAMA network open 4, 3 (2021), e212240–e212240.
  46. Validation of actigraphy for sleep measurement in children with cerebral palsy. Sleep medicine 90 (2022), 65–73.
  47. Multi-horizon predictive models for guiding extracorporeal resource allocation in critically ill COVID-19 patients. Journal of the American Medical Informatics Association (2022).
  48. Distance-Based Detection of Cough, Wheeze, and Breath Sounds on Wearable Devices. Sensors 22, 6 (2022), 2167.
  49. Representation learning for treatment effect estimation from observational data. Advances in Neural Information Processing Systems 31 (2018).
  50. GANITE: Estimation of individualized treatment effects using generative adversarial nets. In International Conference on Learning Representations.
Citations (5)
View on Semantic Scholar

Summary

No one has generated a summary of this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Summarize

Paper to Video (Beta)

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Sign Up to Generate

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up