Abstract
The advancement of Electronic Health Records (EHRs) and machine learning have enabled a data-driven and personalised approach to healthcare. One step in this direction is to uncover patient sub-types with similar disease trajectories in a heterogeneous population. This is especially important in the context of mechanical ventilation in intensive care, where mortality is high and there is no consensus on treatment. In this work, we present a new approach to clustering mechanical ventilation episodes, using a multi-task combination of supervised, self-supervised and unsupervised learning techniques. Our dynamic clustering assignment is explicitly guided to reflect the phenotype, trajectory and outcomes of the patient. Experimentation on a real-world dataset is encouraging, and we hope that we could someday translate this into actionable insights in guiding future clinical research.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A tracheostomy is a procedure designed for long term mechanical ventilation of a patient.
- 2.
Preliminary experiments revealed that k-means were more likely to produce small clusters which lay far away from the rest of the data, because it is more affected by outliers. This made the clustering process less reliable and reproducible.
- 3.
There is a 5 h gap between these predictions, therefore this time difference needs to be removed from the first prediction.
- 4.
This is because younger patients can mask a problem by compensating deceptively well, until they reach a point where the homeostatic mechanisms can no longer cope.
References
D.B. Antcliffe, K.L. Burnham, F. Al-Beidh, S. Santhakumaran, S.J. Brett, C.J. Hinds, D. Ashby, J.C. Knight, A.C. Gordon, Transcriptomic signatures in sepsis and a differential response to steroids. From the VANISH randomized trial. Am. J. Respir. Crit. Care. Med. 199(8), 980ā986 (2019)
T. Bein, S. Grasso, O. Moerer, M. Quintel, C. Guerin, M. Deja, A. Brondani, S. Mehta, The standard of care of patients with ARDS: ventilatory settings and rescue therapies for refractory hypoxemia. Intensiv. Care Med. 42(5), 699ā711 (2016)
S.V. Bhavani, M. Semler, E.T. Qian, P.A. Verhoef, C. Robichaux, M.M. Churpek, C.M. Coopersmith, Development and validation of novel sepsis subphenotypes using trajectories of vital signs. Intensiv. Care Med. 48(11), 1582ā1592 (2022)
I. Chami, A. Gu, V. Chatziafratis, C. RĆ©, From trees to continuous embeddings and back: hyperbolic hierarchical clustering (2020). CoRR arXiv.org/abs/2010.00402
Z. Che, S. Purushotham, K. Cho, D. Sontag, Y. Liu, Recurrent neural networks for multivariate time series with missing values. Sci. Rep. 8(1), 6085 (2018)
E. Choi, M.T. Bahadori, A. Schuetz, W.F. Stewart, J. Sun, Doctor AI: predicting clinical events via recurrent neural networks, in JMLR Workshop and Conference, vol. 56 (2015), pp. 301ā318
G. Corso, R. Ying, M. PĆ”ndy, P. VeliÄkoviÄ, J. Leskovec, P. LiĆ², Neural distance embeddings for biological sequences (2021). https://doi.org/10.48550/ARXIV.2109.09740, https://arxiv.org/abs/2109.09740
K.R. Famous, K. Delucchi, L.B. Ware, K.N. Kangelaris, K.D. Liu, B.T. Thompson, C.S. Calfee, A. Network, Acute respiratory distress syndrome subphenotypes respond differently to randomized fluid management strategy. Am. J. Respir. Crit. Care Med. 195(3), 331ā338 (2017)
A.D.T. Force, V.M. Ranieri, G.D. Rubenfeld, B.T. Thompson, N.D. Ferguson, E. Caldwell, E. Fan, L. Camporota, A.S. Slutsky, Acute respiratory distress syndrome: the Berlin definition. JAMA 307(23), 2526ā2533 (2012). https://jamanetwork.com/journals/jama/articlepdf/1160659/jsc120003_2526_2533.pdf
H. Harutyunyan, H. Khachatrian, D.C. Kale, G. Ver Steeg, A. Galstyan, Multitask learning and benchmarking with clinical time series data. Sci. Data 6(96) (2019)
S. Ioffe, C. Szegedy, Batch normalization: accelerating deep network training by reducing internal covariate shift, in Proceedings of the 32nd International Conference on International Conference on Machine Learning vol. 37, JMLR, ICMLā15 (2015), pp. 448ā456
C. Lee, M. vanĀ der Schaar, Temporal phenotyping using deep predictive clustering of disease progression (2020). arXiv.org/abs/2006.08600
Z.C. Lipton, D.C. Kale, C. Elkan, R.C. Wetzel, Learning to diagnose with LSTM recurrent neural networks (2015). CoRR. arXiv:1511.03677
J. MĆ”ca, O. Jor, M. Holub, P. Sklienka, F. BurÅ”a, M. Burda, V. Janout, P. Å evÄĆk, Past and present ARDS mortality rates: a systematic review. Respir. Care 62(1), 113ā122 (2017). http://rc.rcjournal.com/content/62/1/113.full.pdf
R. Miotto, L. Li, B.A. Kidd, J.T. Dudley, Deep patient: an unsupervised representation to predict the future of patients from the electronic health records. Sci. Rep. 6(1), 26094 (2016)
A. Patel, D.M. Montserrat, C. Bustamante, A. Ioannidis, Hyperbolic geometry-based deep learning methods to produce population trees from genotype data (2022). bioRxiv
J. Poole, C. McDowell, R. Lall, G. Perkins, D.F. McAuley, F. Gao, D. Young, Individual patient data analysis of tidal volumes used in three large randomized control trials involving patients with acute respiratory distress syndrome. BJA: Br. J. Anaesth. 118(4), 570ā575 (2017). http://www.oup/backfile/content_public/journal/bja/118/4/10.1093_bja_aew465/1/aew465.pdf
A. Rajkomar, E. Oren, K. Chen et al., Scalable and accurate deep learning with electronic health records. Nature 1(1), 18 (2018)
E Rocheteau, P LiĆ², S Hyland, Temporal pointwise convolutional networks for length of stay prediction in the intensive care unit, in Proceedings of the Conference on Health, Inference, and Learning, Association for Computing Machinery, New York, NY, USA, CHILā21, (2021), pp. 58ā68
A. Rusanov, P.V. Prado, C. Weng, Unsupervised time-series clustering over lab data for automatic identification of uncontrolled diabetes, in 2016 IEEE International Conference on Healthcare Informatics (ICHI) (2016), pp. 72ā80
S. Sheikhalishahi, V. Balaraman, V. Osmani, Benchmarking machine learning models on eICU critical care dataset (2019). 1910.00964
B. Shickel, T.J. Loftus, L. Adhikari, T. Ozrazgat-Baslanti, A. Bihorac, P. Rashidi, DeepSOFA: a continuous acuity score for critically Ill patients using clinically interpretable deep learning. Sci. Rep. (2019)
H. Song, D. Rajan, J. Thiagarajan, A. Spanias, Attend and diagnose: clinical time series analysis using attention models, in 32nd AAAI Conference on Artificial Intelligence, AAAI 2018 (2018), pp. 4091ā4098
P.J. Thoral, J.M. Peppink, R.H. Driessen, E.J.G. Sijbrands, E.J.O. Kompanje, L. Kaplan, H. Bailey, J. Kesecioglu, M. Cecconi, M. Churpek, G. Clermont, M. vanĀ der Schaar, A. Ercole, A.R.J. Girbes, P.W.G. Elbers, Sharing ICU patient data responsibly under the society of critical care medicine/European Society of Intensive Care Medicine Joint Data Science Collaboration: The Amsterdam University Medical Centers Database (AmsterdamUMCdb) Example. Crit. Care Med. 49(6) (2021)
C. Tong, E. Rocheteau, P. VeliÄkoviÄ, N. Lane, P. LiĆ², Predicting patient outcomes with graph representation learning (Springer International Publishing, Cham, 2022), pp.281ā293
A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, U. Kaiser, I. Polosukhin, Attention is all you need, in Proceedings of the 31st International Conference on Neural Information Processing Systems, Curran Associates Inc., NIPSā17 (2017), pp. 6000ā6010
Y. Wang, Y. Zhao, T.M. Therneau, E.J. Atkinson, A.P. Tafti, N. Zhang, S. Amin, A.H. Limper, S. Khosla, H. Liu, Unsupervised machine learning for the discovery of latent disease clusters and patient subgroups using electronic health records. J. Biomed. Inf. 102, 103364 (2020)
H. YĆØche, G. Dresdner, F. Locatello, M. HĆ¼ser, G. RƤtsch, Neighborhood contrastive learning applied to online patient monitoring, in Proceedings of the 38th International Conference on Machine Learning, PMLR, Proceedings of Machine Learning Research, vol. 139, ed. by M. Meila, T. Zhang (2021), pp. 11964ā11974
X. Zhang, J. Chou, J. Liang, C. Xiao, Y. Zhao, H. Sarva, C. Henchcliffe, F. Wang, Data-driven subtyping of Parkinsonās disease using longitudinal clinical records: a cohort study. Sci. Rep. 9(1), 797 (2019)
Acknowledgements
The authors would like to thank Petar VeliÄkoviÄ, Sophie Xhonneux, Stephanie Hyland and Mihaela van der Schaar for helpful discussions and advice. We also thank the Armstrong Fund, the Frank Edward Elmore Fund, and the School of Clinical Medicine at the University of Cambridge for their generous funding.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
Ā© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Rocheteau, E., Bica, I., LiĆ², P., Ercole, A. (2023). Dynamic Outcomes-Based Clustering of Disease Trajectory in Mechanically Ventilated Patients. In: Shaban-Nejad, A., Michalowski, M., Bianco, S. (eds) Artificial Intelligence for Personalized Medicine. W3PHAI 2023. Studies in Computational Intelligence, vol 1106. Springer, Cham. https://doi.org/10.1007/978-3-031-36938-4_6
Download citation
DOI: https://doi.org/10.1007/978-3-031-36938-4_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-36937-7
Online ISBN: 978-3-031-36938-4
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)