Skip to main content

Hypotensive Episode Prediction in ICUs via Observation Window Splitting

Part of the Lecture Notes in Computer Science book series (LNAI,volume 11053)

Abstract

Hypotension, defined as dangerously low blood pressure, is a significant risk factor in intensive care units (ICUs), which requires a prompt therapeutic intervention. The goal of our research is to predict an impending Hypotensive Episode (HE) by time series analysis of continuously monitored physiological vital signs. Our prognostic model is based on the last Observation Window (OW) at the prediction time. Existing clinical episode prediction studies used a single OW of 5–120 min to extract predictive features, with no significant improvement reported when longer OWs were used. In this work we have developed the In-Window Segmentation (InWiSe) method for time series prediction, which splits a single OW into several sub-windows of equal size. The resulting feature set combines the features extracted from each observation sub-window and then this combined set is used by the Extreme Gradient Boosting (XGBoost) binary classifier to produce an episode prediction model. We evaluate the proposed approach on three retrospective ICU datasets (extracted from MIMIC II, Soroka and Hadassah databases) using cross-validation on each dataset separately, as well as by cross-dataset validation. The results show that InWiSe is superior to existing methods in terms of the area under the ROC curve (AUC).

Keywords

  • Time series analysis
  • Clinical episode prediction
  • Feature extraction
  • Intensive care
  • Patient monitoring

Partially supported by the Cincinnati Children’s Hospital Medical Center; In collaboration with Soroka Medical Center in Beer-Sheva and Hadassah University Hospital, Ein Karem, Jerusalem

This is a preview of subscription content, access via your institution.

Buying options

Chapter
USD   29.95
Price excludes VAT (USA)
  • DOI: 10.1007/978-3-030-10997-4_29
  • Chapter length: 16 pages
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
eBook
USD   84.99
Price excludes VAT (USA)
  • ISBN: 978-3-030-10997-4
  • Instant PDF download
  • Readable on all devices
  • Own it forever
  • Exclusive offer for individuals only
  • Tax calculation will be finalised during checkout
Softcover Book
USD   109.99
Price excludes VAT (USA)
Fig. 1.
Fig. 2.
Fig. 3.

Notes

  1. 1.

    Recently, The MIMIC III waveform database Matched Subset, four times larger than the MIMIC II subset, was published

  2. 2.

    All improvement percentages are in terms of a ratio between the two measures

References

  1. Ghosh, S., et al.: Hypotension risk prediction via sequential contrast patterns of ICU blood pressure. IEEE J. Biomed. Health Inform. 20(5), 1416–1426 (2016)

    CrossRef  Google Scholar 

  2. Sebat, F., et al.: Effect of a rapid response system for patients in shock on time to treatment and mortality during 5 years. Crit. Care Med. 35(11), 2568–2575 (2007)

    CrossRef  Google Scholar 

  3. Cao, H., et al.: Predicting ICU hemodynamic instability using continuous multiparameter trends. In: Engineering in Medicine and Biology Society (EMBS), pp. 3803–3806. IEEE (2008)

    Google Scholar 

  4. Moody, G.B., Lehman, L.W.H.: Predicting acute hypotensive episodes: the 10th annual physionet/computers in cardiology challenge. In: Computers in Cardiology, pp. 541–544. IEEE (2009)

    Google Scholar 

  5. Forkan, A.R.M., et al.: ViSiBiD: a learning model for early discovery and real-time prediction of severe clinical events using vital signs as big data. Comput. Netw. 113, 244–257 (2017)

    CrossRef  Google Scholar 

  6. Kamio, T., et al.: Use of machine-learning approaches to predict clinical deterioration in critically Ill patients: a systematic review. Int. J. Med. Res. Health Sci. 6(6), 1–7 (2017)

    Google Scholar 

  7. Eshelman, L.J., et al.: Development and evaluation of predictive alerts for hemodynamic instability in ICU patients. In: 2008 AMIA Annual Symposium Proceedings, p. 379. American Medical Informatics Association (2008)

    Google Scholar 

  8. Lee, J., Mark, R.G.: An investigation of patterns in hemodynamic data indicative of impending hypotension in intensive care. Biomed. Eng. Online 9(1), 62 (2010)

    CrossRef  Google Scholar 

  9. Saeed, M., et al.: Multiparameter intelligent monitoring in intensive care II (MIMIC-II): a public-access ICU database. Crit. Car. Med. 39(5), 952 (2011)

    CrossRef  Google Scholar 

  10. Chen, X., et al.: Forecasting acute hypotensive episodes in intensive care patients based on a peripheral arterial blood pressure waveform. In: 2009 Computers in Cardiology, pp. 545–548. IEEE (2009)

    Google Scholar 

  11. Saeed, M.: Temporal pattern recognition in multiparameter ICU data, Doctoral dissertation, Massachusetts Institute of Technology (2007)

    Google Scholar 

  12. Saeed, M., Mark, R.: A novel method for the efficient retrieval of similar multiparameter physiologic time series using wavelet-based symbolic representations. In: AMIA Annual Symposium Proceedings, p. 679. American Medical Information Association (2006)

    Google Scholar 

  13. Rocha, T., et al.: Wavelet based time series forecast with application to acute hypotensive episodes prediction. In: Engineering in medicine and biology society (EMBC), pp. 2403–2406. IEEE (2010)

    Google Scholar 

  14. Ghosh, S., et al.: Septic shock prediction for ICU patients via coupled HMM walking on sequential contrast patterns. J. Biomed. Info. 66, 19–31 (2017)

    CrossRef  Google Scholar 

  15. Lee, J., Mark, R.G.: A hypotensive episode predictor for intensive care based on heart rate and blood pressure time series. In: Computing in Cardiology, pp. 81–84. IEEE (2010)

    Google Scholar 

  16. Rocha, T., et al.: Prediction of acute hypotensive episodes by means of neural network multi-models. Comp. Biol. Med. 41(10), 881–890 (2011)

    CrossRef  Google Scholar 

  17. Goldberger, A.L., et al.: PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. Circulation 101(23), e215–e220 (2000)

    CrossRef  Google Scholar 

  18. The MIMIC II Waveform Database Matched Subset (Physionet Database). https://physionet.org/physiobank/database/mimic2wdb/matched/

  19. Lee, Y.-L., Juan, D.-C., Tseng, X.-A., Chen, Y.-T., Chang, S.-C.: DC-Prophet: predicting catastrophic machine failures in DataCentre. In: Altun, Y., et al. (eds.) Machine Learning and Knowledge Discovery in Databases. LNCS, vol. 10536, pp. 64–76. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-71273-4_6

    CrossRef  Google Scholar 

  20. Chen, T., Guestrin, C.: XGBoost: a scalable tree boosting system. In: 22nd ACM SIGKDD International Conference, pp. 785–794. ACM (2016)

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Mark Last .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and Permissions

Copyright information

© 2019 Springer Nature Switzerland AG

About this paper

Verify currency and authenticity via CrossMark

Cite this paper

Tsur, E., Last, M., Garcia, V.F., Udassin, R., Klein, M., Brotfain, E. (2019). Hypotensive Episode Prediction in ICUs via Observation Window Splitting. In: , et al. Machine Learning and Knowledge Discovery in Databases. ECML PKDD 2018. Lecture Notes in Computer Science(), vol 11053. Springer, Cham. https://doi.org/10.1007/978-3-030-10997-4_29

Download citation

  • DOI: https://doi.org/10.1007/978-3-030-10997-4_29

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-030-10996-7

  • Online ISBN: 978-3-030-10997-4

  • eBook Packages: Computer ScienceComputer Science (R0)