skip to main content
research-article

Attention-Based Deep Recurrent Model for Survival Prediction

Published:14 September 2021Publication History
Skip Abstract Section

Abstract

Survival analysis exhibits profound effects on health service management. Traditional approaches for survival analysis have a pre-assumption on the time-to-event probability distribution and seldom consider sequential visits of patients on medical facilities. Although recent studies leverage the merits of deep learning techniques to capture non-linear features and long-term dependencies within multiple visits for survival analysis, the lack of interpretability prevents deep learning models from being applied to clinical practice. To address this challenge, this article proposes a novel attention-based deep recurrent model, named AttenSurv, for clinical survival analysis. Specifically, a global attention mechanism is proposed to extract essential/critical risk factors for interpretability improvement. Thereafter, Bi-directional Long Short-Term Memory is employed to capture the long-term dependency on data from a series of visits of patients. To further improve both the prediction performance and the interpretability of the proposed model, we propose another model, named GNNAttenSurv, by incorporating a graph neural network into AttenSurv, to extract the latent correlations between risk factors. We validated our solution on three public follow-up datasets and two electronic health record datasets. The results demonstrated that our proposed models yielded consistent improvement compared to the state-of-the-art baselines on survival analysis.

References

  1. Ahmed M. Alaa and Mihaela van der Schaar. 2017. Deep multi-task Gaussian processes for survival analysis with competing risks. In Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS’17). 2326–2334. Google ScholarGoogle ScholarDigital LibraryDigital Library
  2. Peipei Chen, Wei Dong, Xudong Lu, Uzay Kaymak, Kunlun He, and Zhengxing Huang. 2019. Deep representation learning for individualized treatment effect estimation using electronic health records. Journal of Biomedical Informatics 100 (Oct. 2019), Article 103303, 11 pages.Google ScholarGoogle Scholar
  3. David Cox. 1972. Regression models and life-tables. Journal of the Royal Statistical Society: Series B (Methodological) 34, 2 (March 1972), 187–220.Google ScholarGoogle Scholar
  4. Christina Curtis, P. Shah Sohrab, Suet-Feung Chin, Turashvili Gulisa, Oscar M. Rueda, Mark J. Dunning, Doug Speed, et al. 2012. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 7403 (April 2012), 346–352.Google ScholarGoogle Scholar
  5. Huilong Duan, Zhoujian Sun, Wei Dong, He Kunlun, and Huang Zhengxing. 2020. On clinical event prediction in patient treatment trajectory using longitudinal electronic health records. IEEE Journal of Biomedical and Health Informatics 24, 7 (July 2020), 2053–2063.Google ScholarGoogle Scholar
  6. Tamara Fernández, Nicolás Rivera, and Yee Whye Teh. 2016. Gaussian processes for survival analysis. In Proceedings of the 30st International Conference on Neural Information Processing Systems (NIPS’16). 5021–5029. Google ScholarGoogle ScholarDigital LibraryDigital Library
  7. Edmund A. Gehan. 1965. A generalized Wilcoxon test for comparing arbitrarily singly-censored samples. Biometrika 52, 1–2 (June 1965), 203–224.Google ScholarGoogle ScholarCross RefCross Ref
  8. Eva Gerdts, Kristian Wachtell, Per Omvik, Jan Erik Otterstad, Oikarinen Lasse, Boman Kurt, Dahlöf Björn, and Richard B. Devereux. 2006. Left atrial size and risk of major cardiovascular events during antihypertensive treatment: Losartan intervention for endpoint reduction in hypertension trial. Hypertension 49, 2 (July 2006), 311–316.Google ScholarGoogle Scholar
  9. Richard Gill and Martin Schumacher. 1987. A simple test of the proportional hazards assumption. Biometrika 74, 2 (June 1987), 289–300.Google ScholarGoogle Scholar
  10. Eleonora Giunchiglia, Anton Nemchenko, and Mihaela van der Schaar. 2018. RNN-SURV: A deep recurrent model for survival analysis. In Proceedings of the 27th International Conference on Artificial Neural Networks (ICANN’18). 23–32.Google ScholarGoogle Scholar
  11. Kumiko Hamano, Ikue Nakadaira, Jun Suzuki, and Gonai Megumi. 2014. N-terminal fragment of probrain natriuretic peptide is associated with diabetes microvascular complications in type 2 diabetes. Vascular Health and Risk Management 10 (Oct. 2014), 585–589.Google ScholarGoogle Scholar
  12. Zhengxing Huang and Wei Dong. 2019. Adversarial MACE prediction after acute coronary syndrome using electronic health records. IEEE Journal of Biomedical and Health Informatics 23, 5 (Sept. 2019), 2117–2126.Google ScholarGoogle Scholar
  13. Zhengxing Huang, Wei Dong, Huilong Duan, and Liu Jiquan. 2017. A regularized deep learning approach for clinical risk prediction of acute coronary syndrome using electronic health records. IEEE Transactions on Biomedical Engineering 65, 5 (July 2017), 956–968.Google ScholarGoogle Scholar
  14. Hemant Ishwaran, Udaya B. Kogalur, Eugene H. Blackstone, and Michael S. Lauer. 2008. Random survival forests. Annals of Applied Statistics 2, 3 (March 2008), 841–860.Google ScholarGoogle ScholarCross RefCross Ref
  15. Daniel Jarrett, Jinsung Yoon, and Mihaela van der Schaar. 2021. Dynamic prediction in clinical survival analysis using temporal convolutional networks. IEEE Journal of Biomedical and Health Informatics 24, 2 (Feb. 2021), 424–436.Google ScholarGoogle Scholar
  16. Alistair E. W. Johnson, Tom J. Pollard, Lu Shen, H. Lehman Li-Wei, Mengling Feng, Mohammad Ghassemi, Benjamin Moody, Peter Szolovits, Leo Anthony Celi, and Roger G. Mark. 2016. MIMIC-III, a freely accessible critical care database. Scientific Sata 3 (May 2016), Article 160035, 9 pages.Google ScholarGoogle Scholar
  17. Edward L. Kaplan and Paul Meier. 1958. Nonparametric estimation from incomplete observations. Journal of the American Statistical Association 53, 282 (June 1958), 457–481.Google ScholarGoogle ScholarCross RefCross Ref
  18. Jared Katzman, Uri Shaham, Alexander Cloninger, Bates Jonathan, Jiang Tingting, and Kluger Yuval. 2018. DeepSurv: Personalized treatment recommender system using a Cox proportional hazards deep neural network. BMC Medical Research Methodology 18, 24 (Feb. 2018), 12.Google ScholarGoogle ScholarCross RefCross Ref
  19. Kunihiro Kinjo, Hiroshi Sato, Hideyuki Sato, Ohnishi Yozo, Hishida Eiji, Nakatani Daisaku, Mizuno Hiroya, et al. 2003. Prognostic significance of atrial fibrillation/atrial flutter in patients with acute myocardial infarction treated with percutaneous coronary intervention. American Journal of Cardiology 92, 10 (Nov. 2003), 1150–1154.Google ScholarGoogle Scholar
  20. William A. Knaus, Frank E. Harrell, Joanne Lynn, L. Goldman, R. S. Phillips, A. F. Connors Jr., N. V. Dawson, et al. 1995. The SUPPORT prognostic model: Objective estimates of survival for seriously ill hospitalized adults. Annals of Internal Medicine 122, 3 (Feb. 1995), 191–203.Google ScholarGoogle Scholar
  21. Harlan M. Krumholz, Eugene M. Parent, Nora Tu, V. Vaccarino, Y. Wang, M. J. Radford, and J. Hennen. 1997. Readmission after hospitalization for congestive heart failure among Medicare beneficiaries. Archives of Internal Medicine 157, 1 (Jan. 1997), 99–104.Google ScholarGoogle Scholar
  22. Harlan M. Krumholz, Ya-Ting Chen, Viola Vaccarino, Yun Wang, Martha J. Radford, W. David Bradford, and Ralph I. Horwitz. 2000. Correlates and impact on outcomes of worsening renal function in patients >= 65 years of age with heart failure. American Journal of Cardiology 85, 9 (May 2000), 1110–1113.Google ScholarGoogle Scholar
  23. Changhee Lee, Jinsung Yoon, and Mihaela Schaar. 2020. Dynamic-DeepHit: A deep learning approach for dynamic survival analysis with competing risks based on longitudinal data. IEEE Transactions on Biomedical Engineering 67, 1 (Jan. 2020), 122–133.Google ScholarGoogle Scholar
  24. Changhee Lee, William R. Zame, Jinsung Yoon, and Mihaela van der Schaar. 2018. DeepHit: A deep learning approach to survival analysis with competing risks. In Proceedings of 32nd AAAI Conference on Artificial Intelligence (AAAI’18). 2314–2321.Google ScholarGoogle Scholar
  25. KyuHa Lee, Sounak Chakraborty, and Jianguo Sun. 2011. Bayesian variable selection in semiparametric proportional hazards model for high dimensional survival data. International Journal of Biostatistics 7, 1 (April 2011), 1–32.Google ScholarGoogle ScholarCross RefCross Ref
  26. Yujia Li, Daniel Tarlow, Brockschmidt Marc, and Zemel Richard. 2015. Gated graph sequence neural networks. arXiv:1511.05493.Google ScholarGoogle Scholar
  27. Zekun Li, Zeyu Cui, Shu Wu, Xiaoyu Zhang, and Liang Wang. 2019. Fi-GNN: Modeling feature interactions via graph neural networks for CTR prediction. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management (CIKM’19). 539–548. Google ScholarGoogle ScholarDigital LibraryDigital Library
  28. Xian Liu. 2012. Survival Analysis: Models and Applications. John Wiley & Sons, Hoboken, NJ.Google ScholarGoogle ScholarCross RefCross Ref
  29. Nathan Mantel. 1966. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemotherapy Reports 50, 3 (March 1966), 163–170.Google ScholarGoogle Scholar
  30. Frederick A. Masoudi, Edward P. Havranek, Grace Smith, Ronald H. Fish, John F. Steiner, Diana L. Ordin, and Harlan M. Krumholz. 2003. Gender, age, and heart failure with preserved left ventricular systolic function. Journal of the American College of Cardiology 41, 2 (Jan. 2003), 217–223.Google ScholarGoogle Scholar
  31. Fen Miao, Yun-Peng Cai, Yuan-Ting Zhang, and Chun-Yue Li. 2015. Is random survival forest an alternative to Cox proportional model on predicting cardiovascular disease?Medical and Biological Engineering and Computing 1, 1 (Sept. 2015), 740–743.Google ScholarGoogle Scholar
  32. Freeman Michael and Zeegers Maurice (Eds.). 2016. Forensic Epidemiology. Academic Press, Amsterdam, Netherlands.Google ScholarGoogle Scholar
  33. Milad Zafar Nezhad, Najibesadat Sadati, Kai Yang, and Dongxiao Zhu. 2019. A deep active survival analysis approach for precision treatment recommendations: Application of prostate cancer. Expert Systems with Applications 115, 1 (Jan. 2019), 16–26.Google ScholarGoogle Scholar
  34. Stuart J. Pocock, Cono A. Ariti, John J. V. McMurray, Aldo Maggioni, Lars Køber, Iain B. Squire, Karl Swedberg, et al. 2012. Predicting survival in heart failure: A risk score based on 39372 patients from 30 studies. European Heart Journal 34, 19 (Oct. 2012), 1404–1413.Google ScholarGoogle Scholar
  35. Kan Ren, Jiarui Qin, Lei Zheng, Zhengyu Yang, Weinan Zhang, Qiu Lin, and Yu Yong. 2019. Deep recurrent survival analysis. In Proceedings of the 33rd AAAI Conference on Artificial Intelligence (AAAI’19). 4798–4805.Google ScholarGoogle Scholar
  36. Mike Schuster and Kuldip K. Paliwal. 1997. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing 45, 11 (Nov. 1997), 2673–2681. Google ScholarGoogle ScholarDigital LibraryDigital Library
  37. Sama Shrestha, Bret D. Elderd, and Vanja Dukic. 2019. Bayesian-based survival analysis: Inferring time to death in host-pathogen interactions. Environmental and Ecological Statistics 26, 1 (Feb. 2019), 17–45.Google ScholarGoogle Scholar
  38. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS’17). 6000–6010. Google ScholarGoogle ScholarDigital LibraryDigital Library
  39. Hong Wang and Gang Li. 2017. A selective review on random survival forests for high dimensional data. Quantitative Bioscience 36, 2 (2017), 85–96.Google ScholarGoogle Scholar
  40. David W. Hosmer, Stanley Lemeshow, and Susanne May. 2008. Applied Survival Analysis. John Wiley & Sons, Hoboken, NJ.Google ScholarGoogle Scholar
  41. Shannon Wongvibulsin, Katherine C. Wu, and Scott L. Zeger. 2019. Clinical risk prediction with random forests for survival, longitudinal, and multivariate (RF-SLAM) data analysis. BMC Medical Research Methodology 20, 1 (Dec. 2019), 1–14.Google ScholarGoogle Scholar
  42. Wojciech Zaremba, Ilya Sutskever, and Oriol Vinyals. 2014. Recurrent neural network regularization. arXiv:1409.2329.Google ScholarGoogle Scholar
  43. Xinliang Zhu, Jiawen Yao, and Junzhou Huang. 2016. Deep convolutional neural network for survival analysis with pathological images. In Proceedings of the 2016 IEEE International Conference on Bioinformatics and Biomedicine (BIBM’16). 544–547.Google ScholarGoogle Scholar
  44. Carlos Álvarez Zurro, Antonio Planas Roca, Enrique Alday Muñoz, Lorena Vega Piris, Fernando Ramasco Rueda, and Rosa Méndez Hernández. 2016. High levels of preoperative and postoperative N terminal B-type natriuretic propeptide influence mortality and cardiovascular complications after noncardiac surgery: A prospective cohort study. European Journal of Anaesthesiology 33, 6 (June 2016), 444–449.Google ScholarGoogle Scholar

Index Terms

  1. Attention-Based Deep Recurrent Model for Survival Prediction

          Recommendations

          Comments

          Login options

          Check if you have access through your login credentials or your institution to get full access on this article.

          Sign in

          Full Access

          PDF Format

          View or Download as a PDF file.

          PDF

          eReader

          View online with eReader.

          eReader

          HTML Format

          View this article in HTML Format .

          View HTML Format
          About Cookies On This Site

          We use cookies to ensure that we give you the best experience on our website.

          Learn more

          Got it!