Advertisement

Using a LSTM-RNN Based Deep Learning Framework for ICU Mortality Prediction

  • Hanzhong Zheng
  • Dejia ShiEmail author
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11242)

Abstract

In Intensive Care Units (ICU), the machine learning technique has been widely used in ICU patient data. A mortality risky model can provide assessment on patients’ current and when the disease may worsen. The prediction of mortality outcomes even intervenes doctor’s decision making on patient’s treatment. Based on the patient’s condition, a timely intervention treatment is adopted to prevent the patient’s condition gets worse. However, the common major challenges in ICU patient data are irregular data sampling and missing variables values. In this paper, we used a statistical approach to preprocess the data. We introduced a data imputation method based on Gaussian process and proposed a deep learning technology using LSTM-RUN that emphasizes on long time dependency relation inside the patient data records to predict the probability of patient’s mortality in ICU. The experiment results show that LSTM improved the mortality prediction accuracy than base RNN using the new statistical imputation method for handling missing data problem.

Keywords

Deep learning Recurrent neural network Mortality prediction Gaussian process 

References

  1. 1.
    Baytas, I.M., Xiao, C., Zhang, X., Wang, F., Jain, A.K., Zhou, J.: Patient subtyping via time-aware LSTM networks. In: Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2017, pp. 65–74. ACM, New York (2017).  https://doi.org/10.1145/3097983.3097997
  2. 2.
    Che, Z., Kale, D., Li, W., Bahadori, M.T., Liu, Y.: Deep computational phenotyping. In: Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2015, pp. 507–516. ACM, New York (2015).  https://doi.org/10.1145/2783258.2783365
  3. 3.
    Deligiannis, P., Loidl, H.W., Kouidi, E.: Improving the diagnosis of mild hypertrophic cardiomyopathy with mapreduce. In: Proceedings of Third International Workshop on MapReduce and Its Applications Date, pp. 41–48, MapReduce 2012. ACM, New York, NY, USA (2012).  https://doi.org/10.1145/2287016.2287025
  4. 4.
    Harutyunyan, H., Khachatrian, H., Kale, D., Galstyan, A.: Multitask learning and benchmarking with clinical time series data, March 2017Google Scholar
  5. 5.
    Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997).  https://doi.org/10.1162/neco.1997.9.8.1735CrossRefGoogle Scholar
  6. 6.
    Hussain, A.: Machine learning approaches for extracting genetic medical data information. In: Proceedings of the Second International Conference on Internet of Things, Data and Cloud Computing, ICC 2017, p. 1. ACM, New York (2017).  https://doi.org/10.1145/3018896.3066906

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.Department of Computer ScienceUniversity of PittsburghPittsburghUSA
  2. 2.Key Laboratory of Hunan Province for New Retail Virtual Reality TechnologyHunan University of CommerceChangshaChina

Personalised recommendations