Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Clinical Research Article
  • Published:

Detection and quantitative analysis of patient-ventilator interactions in ventilated infants by deep learning networks

Pediatric Research (2024)Cite this article

Abstract

Background

The study of patient-ventilator interactions (PVI) in mechanically ventilated neonates is limited by the lack of unified PVI definitions and tools to perform large scale analyses.

Methods

An observational study was conducted in 23 babies randomly selected from 170 neonates who were ventilated with SIPPV-VG, SIMV-VG or PSV-VG mode for at least 12 h. 500 breaths were randomly selected and manually annotated from each recording to train convolutional neural network (CNN) models for PVI classification.

Results

The average asynchrony index (AI) over all recordings was 52.5%. The most frequently occurring PVIs included expiratory work (median: 28.4%, interquartile range: 23.2–40.2%), late cycling (7.6%, 2.8–10.2%), failed triggering (4.6%, 1.2–6.2%) and late triggering (4.4%, 2.8–7.4%). Approximately 25% of breaths with a PVI had two or more PVIs occurring simultaneously. Binary CNN classifiers were developed for PVIs affecting ≥1% of all breaths (n = 7) and they achieved F1 scores of >0.9 on the test set except for early triggering where it was 0.809.

Conclusions

PVIs occur frequently in neonates undergoing conventional mechanical ventilation with a significant proportion of breaths containing multiple PVIs. We have developed computational models for seven different PVIs to facilitate automated detection and further evaluation of their clinical significance in neonates.

Impact

  • The study of patient-ventilator interactions (PVI) in mechanically ventilated neonates is limited by the lack of unified PVI definitions and tools to perform large scale analyses.

  • By adapting a recent taxonomy of PVI definitions in adults, we have manually annotated neonatal ventilator waveforms to determine prevalence and co-occurrence of neonatal PVIs.

  • We have also developed binary deep learning classifiers for common PVIs to facilitate their automatic detection and quantification.

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

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Examples of pressure (upper panels) and flow (lower panels) waveforms of patient ventilator interactions (PVIs) studied in this paper.
Fig. 2: Workflow showing development of deep learning classifiers to detect specific patient-ventilator interactions.

Similar content being viewed by others

Data availability

The computer code of the deep learning models and their analysis can be freely accessed at https://github.com/chongtwd/Detection-and-quantitative-analysis-of-patient-ventilator-interactions-in-ventilated-neonates. The ventilator data used to train and evaluate the models are not publicly available due to ethical concerns but are available from the corresponding author on reasonable request and subject to favorable ethical opinion.

References

  1. Brown, M. K. & DiBlasi, R. M. Mechanical ventilation of the premature neonate. Respir. Care 56, 1298–1311 (2011).

    Article  PubMed  Google Scholar 

  2. Norman, M., Jonsson, B., Wallström, L. & Sindelar, R. Respiratory support of infants born at 22–24 weeks of gestational age. Semin. Fetal Neonatal Med. 27, 101328 (2022).

    Article  PubMed  Google Scholar 

  3. Miller, J. D. & Carlo, W. A. Pulmonary complications of mechanical ventilation in neonates. Clin. Perinatol. 35, 273–281 (2008).

    Article  PubMed  Google Scholar 

  4. Baker, C. D. Chronic respiratory failure in bronchopulmonary dysplasia. Pediatr. Pulmonol. 56, 3490–3498 (2021).

    Article  PubMed  Google Scholar 

  5. DeMauro, S. B. Neurodevelopmental outcomes of infants with bronchopulmonary dysplasia. Pediatr. Pulmonol. 56, 3509–3517 (2021).

    Article  PubMed  Google Scholar 

  6. Schmalisch, G. Basic principles of respiratory function monitoring in ventilated newborns: a review. Paediatr. Respir. Rev. 20, 76–82 (2016).

    PubMed  Google Scholar 

  7. Beck, J. & Sinderby, C. Neurally adjusted ventilatory assist in newborns. Clin. Perinatol. 48, 783–811 (2021).

    Article  PubMed  Google Scholar 

  8. Hummler, H. & Schulze, A. New and alternative modes of mechanical ventilation in neonates. Semin. Fetal Neonatal Med. 14, 42–48 (2009).

    Article  PubMed  Google Scholar 

  9. van Kaam, A. H. et al. Modes and strategies for providing conventional mechanical ventilation in neonates. Pediatr. Res. 90, 957–962 (2021).

    Article  PubMed  Google Scholar 

  10. Mammel, M. C. & Donn, S. M. Real-time pulmonary graphics. Semin. Fetal Neonatal Med. 20, 181–191 (2015).

    Article  PubMed  Google Scholar 

  11. Crooke, P. S., Head, J. D. & Marini, J. J. A general two-compartment model for mechanical ventilation. Math. Comp. Model. 24, 1–18 (1996).

    Article  Google Scholar 

  12. Bhutani, V. K., Sivieri, E. M., Abbasi, S. & Shaffer, T. H. Evaluation of neonatal pulmonary mechanics and energetics: a two factor least mean square analysis. Pediatr. Pulmonol. 4, 150–158 (1988).

    Article  PubMed  CAS  Google Scholar 

  13. Nilsestuen, J. O. & Hargett, K. D. Using ventilator graphics to identify patient-ventilator asynchrony. Respir. Care 50, 202–234 (2005).

    PubMed  Google Scholar 

  14. Mirabella, L. et al. Patient-ventilator asynchronies: clinical implications and practical solutions. Respir. Care 65, 1751–1766 (2020).

    Article  PubMed  Google Scholar 

  15. Blanch, L. et al. Asynchronies during mechanical ventilation are associated with mortality. Intensive Care Med. 41, 633–641 (2015).

    Article  PubMed  Google Scholar 

  16. Kyo, M. et al. Patient–ventilator asynchrony, impact on clinical outcomes and effectiveness of interventions: a systematic review and meta-analysis. J. Intensive Care 9, 50 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  17. Mortamet, G. et al. Patient–ventilator asynchrony during conventional mechanical ventilation in children. Ann. Intensive Care 7, 122 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Blokpoel, R. G. T., Burgerhof, J. G. M., Markhorst, D. G. & Kneyber, M. C. J. Trends in pediatric patient-ventilator asynchrony during invasive mechanical ventilation. Pediatr. Crit. Care Med. 22, 993–997 (2021).

    Article  PubMed  Google Scholar 

  19. Greenough, A., Morley, C. & Davis, J. Interaction of spontaneous respiration with artificial ventilation in preterm babies. J. Pediatr. 103, 769–773 (1983).

    Article  PubMed  CAS  Google Scholar 

  20. McCallion, N., Lau, R., Dargaville, P. A. & Morley, C. J. Volume guarantee ventilation, interrupted expiration, and expiratory braking. Arch. Dis. Child. 90, 865–870 (2005).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  21. Mally, P. V., Beck, J., Sinderby, C., Caprio, M. & Bailey, S. M. Neural breathing pattern and patient-ventilator interaction during neurally adjusted ventilatory assist and conventional ventilation in newborns. Pediatr. Crit. Care Med. 19, 48–55 (2018).

    Article  PubMed  Google Scholar 

  22. Bignall, S., Dixon, P., Quinn, C. & Kitney, R. Monitoring interactions between spontaneous respiration and mechanical inflations in preterm neonates. Crit. Care Med. 25, 545–553 (1997).

    Article  PubMed  CAS  Google Scholar 

  23. Adams, J. Y. et al. Development and validation of a multi-algorithm analytic platform to detect off-target mechanical ventilation. Sci. Rep. 7, 14980 (2017).

    Article  PubMed  PubMed Central  ADS  Google Scholar 

  24. Blanch, L. et al. Validation of the Better Care® system to detect ineffective efforts during expiration in mechanically ventilated patients: a pilot study. Intensive Care Med. 38, 772–780 (2012).

    Article  PubMed  Google Scholar 

  25. Rehm, G. et al. Creation of a robust and generalizable machine learning classifier for patient ventilator asynchrony. Methods Inf. Med. 57, 208–219 (2018).

    Article  PubMed  Google Scholar 

  26. Casagrande, A. et al. An effective pressure–flow characterization of respiratory asynchronies in mechanical ventilation. J. Clin. Monit. Comput. 35, 289–296 (2020).

    Article  PubMed  Google Scholar 

  27. Gholami, B. et al. Replicating human expertise of mechanical ventilation waveform analysis in detecting patient-ventilator cycling asynchrony using machine learning. Comput. Biol. Med. 97, 137–144 (2018).

    Article  PubMed  Google Scholar 

  28. Bakkes, T. H. G. F., Montree, R. J. H., Mischi, M., Mojoli, F. & Turco, S. A machine learning method for automatic detection and classification of patient-ventilator asynchrony. In: 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). IEEE (2020). Accessed on 7th June, 2022, from https://doi.org/10.1109/embc44109.2020.9175796

  29. Zhang, L. et al. Detection of patient-ventilator asynchrony from mechanical ventilation waveforms using a two-layer long short-term memory neural network. Comp. Biol. Med. 120, 103721 (2020).

    Article  Google Scholar 

  30. Rusconi, F. et al. Reference values for respiratory rate in the first 3 years of life. Pediatrics 94, 350–355 (1994).

    Article  PubMed  CAS  Google Scholar 

  31. Numa, A. H. & Newth, C. J. Anatomic dead space in infants and children. J. Appl. Physiol. 80, 1485–1489 (1996).

    Article  PubMed  CAS  Google Scholar 

  32. Mireles-Cabodevila, E., Siuba, M. T. & Chatburn, R. L. A taxonomy for patient-ventilator interactions and a method to read ventilator waveforms. Respir. Care 67, 129–148 (2021).

    Article  PubMed  Google Scholar 

  33. Chong, D., Morley, C. J. & Belteki, G. Computational analysis of neonatal ventilator waveforms and loops. Pediatr. Res. 89, 1432–1441 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

  34. Thille, A. W., Rodriguez, P., Cabello, B., Lellouche, F. & Brochard, L. Patient-ventilator asynchrony during assisted mechanical ventilation. Intensive Care Med. 32, 1515–1522 (2006).

    Article  PubMed  Google Scholar 

  35. Glorot, X., Bordes, A. & Bengio, Y. Deep sparse rectifier neural networks”, In Int. Conf. Artificial Intelligence and Statistics, 315–323 (2011).

  36. Kingma, D., & Adam, J. B. A Method for Stochastic Optimization. (2014) Retrieved on 3rd August 2022, from https://arxiv.org/abs/1412.6980.

  37. Wikipedia contributors. F-score. In Wikipedia, The Free Encyclopedia. Retrieved on 19th April 2023, from https://en.wikipedia.org/w/index.php?title=F-score&oldid=1148225663

  38. de Waal, C. G., van Leuteren, R. W., de Jongh, F. H., van Kaam, A. H. & Hutten, G. J. Patient-ventilator asynchrony in preterm infants on nasal intermittent positive pressure ventilation. Arch. Dis. Child Fetal Neonatal Ed. 104, F280–F284 (2019).

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank Thomas Krüger, Kreske Brunckhorst and the engineers of Dräger Medical for their advice and the program exporting data from the ventilator. We thank to Lakshana Gunathilagan for performing some of the manual waveform annotation. We thank to Professor Colin Morley and Dr Amanda Ogilvy-Stuart for her advice and comments on the manuscript. G.B. is a consultant to Vyaire Medical (Mettawa, IL, US) and Dräger Medical (Lübeck, Germany).

Funding

This project was indirectly supported by the National Institute of Health and Care Research, UK (NIHR).

Author information

Authors and Affiliations

  1. Neonatal Intensive Care Unit, The Rosie Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

    David Chong & Gusztav Belteki

Authors
  1. David Chong
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Gusztav Belteki
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

DC contributed to the conceptual design of the study, performed the manual annotation, analyzed the data, developed the classifiers, and contributed to the writing of the manuscript. GB contributed to the conceptual design of the study, helped with interpretation of results, and contributed to the writing of the manuscript. All authors reviewed and approved the final manuscript.

Corresponding author

Correspondence to Gusztav Belteki.

Ethics declarations

Competing interests

The authors declare no competing interests.

Informed consent

Informed consent was obtained from parents in writing.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chong, D., Belteki, G. Detection and quantitative analysis of patient-ventilator interactions in ventilated infants by deep learning networks. Pediatr Res (2024). https://doi.org/10.1038/s41390-024-03064-z

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1038/s41390-024-03064-z

Search

Advanced search

Quick links