Online apnea–bradycardia detection based on hidden semi-Markov models

  • Miguel Altuve
  • Guy Carrault
  • Alain Beuchée
  • Patrick Pladys
  • Alfredo I. Hernández
Original Article

Abstract

In this paper, we propose a new online apnea–bradycardia detection scheme that takes into account not only the instantaneous values of time series, but also their temporal evolution. The detector is based on a set of hidden semi-Markov models, representing the temporal evolution of beat-to-beat interval (RR interval) time series. A preprocessing step, including quantization and delayed version of the observation vector, is also proposed to maximize detection performance. This approach is quantitatively evaluated through simulated and real signals, the latter being acquired in neonatal intensive care units (NICU). Compared to two conventional detectors used in NICU, our best detector shows an improvement on average of around 15 % in sensitivity and 7 % in specificity. Furthermore, a reduced detection delay of approximately 2 s is also observed with respect to conventional detectors.

Keywords

Time series analysis Hidden semi-Markov models Data mining Electrocardiogram Apnea–bradycardia 

References

  1. 1.
    Ahlborn V, Bohnhorst B, Peter C, Poets C (2000) False alarms in very low birthweight infants: comparison between three intensive care monitoring systems. Acta Paediatr 89(5):571–576PubMedCrossRefGoogle Scholar
  2. 2.
    Altuve M, Carrault G, Beuchee A, Flamand C, Pladys P, Hernandez AI (2012) Comparing hidden markov model and hidden semi-Markov model based detectors of apnea–bradycardia episodes in preterm infants. In: Computing in cardiology (CinC), Krakow, pp 389–392. Google Scholar
  3. 3.
    Altuve M, Carrault G, Cruz J, Beuchee A, Pladys P, Hernandez A (2011) Multivariate ecg analysis for apnoea-bradycardia detection and characterisation in preterm infants. Int J Biomed Eng Technol 5(2):247–265CrossRefGoogle Scholar
  4. 4.
    Beuchée A (2005) Intérêt de l’analyse de la variabilité du rythme cardiaque en néonatalogie. Comportement des systèmes de régulation cardiovasculaire dans le syndrome apnée/bradycardie du nouveau-né. Ph.D. thesis, Université de Rennes 1Google Scholar
  5. 5.
    Bishop C (1995) Neural networks for pattern recognition. Oxford university press, OxfordGoogle Scholar
  6. 6.
    Bitan Y, Meyer J, Shinar D, Zmora E (2000) Staff actions and alarms in a neonatal intensive care unit. In: Proceedings of the human factors and ergonomics society annual meeting, vol 44. SAGE Publications, pp 17–20Google Scholar
  7. 7.
    Bitan Y, Meyer J, Shinar D, Zmora E (2004) Nurses’ reactions to alarms in a neonatal intensive care unit. Cognit Technol Work 6(4):239–246CrossRefGoogle Scholar
  8. 8.
    Bohnhorst B, Peter CS, Poets CF (2000) Pulse oximeters’ reliability in detecting hypoxemia and bradycardia: comparison between a conventional and two new generation oximeters. Crit Care Med 28(5):1565–1568PubMedCrossRefGoogle Scholar
  9. 9.
    Brockmann PE, Wiechers C, Pantalitschka T, Diebold J, Vagedes J, Poets CF (2013) Under-recognition of alarms in a neonatal intensive care unit. Arch Dis Child Ed 98(6):F524–F527. doi:10.1136/archdischild-2012-303369 CrossRefGoogle Scholar
  10. 10.
    Chambrin MC et al (2001) Alarms in the intensive care unit: how can the number of false alarms be reduced? Crit Care 5(4):184–188CrossRefGoogle Scholar
  11. 11.
    Chatow U, Davidson S, Reichman B, Akselrod S (1995) Development and maturation of the autonomic nervous system in premature and full-term infants using spectral analysis of heart rate fluctuations. Pediatr Res 37(3):294PubMedCrossRefGoogle Scholar
  12. 12.
    Cruz J, Hernández A, Wong S, Carrault G, Beuchee A (2006) Algorithm fusion for the early detection of apnea–bradycardia in preterm infants. In: Computers in Cardiology, Valencia, pp 473–476. Google Scholar
  13. 13.
    Cvach M (2012) Monitor alarm fatigue: an integrative review. Biomed Instrum Technol 46(4):268–277CrossRefGoogle Scholar
  14. 14.
    Derrode S, Pieczynski W (2004) Signal and image segmentation using pairwise Markov chains. In: IEEE Trans Signal Process 52(9):2477–2489CrossRefGoogle Scholar
  15. 15.
    Dumont J (2008) Fouille de dynamiques multivariées, application à des données temporelles en cardiologie. Ph.D. thesis, Université de Rennes 1Google Scholar
  16. 16.
    Dumont J, Carrault G, Gomis P, Wagner G, Hernández A (2009) Detection of myocardial ischemia with hidden semi-Markovian models. Comput Cardiol 2009:121–124Google Scholar
  17. 17.
    Dumont J, Hernández AI, Carrault G (2010) Improving ecg beats delineation with an evolutionary optimization process. IEEE Trans Biomed Eng 57(3):607–15CrossRefGoogle Scholar
  18. 18.
    Dumont J, Hernandez AI, Fleureau J, Carrault G (2008) Modelling temporal evolution of cardiac electrophysiological features using hidden semi-Markov models. In: Engineering in medicine and biology society, 2008. EMBS 2008. 30th annual international conference of the IEEE. IEEE, pp 165–168Google Scholar
  19. 19.
    Fahrmeir L (1992) Posterior mode estimation by extended Kalman filtering for multivariate dynamic generalized linear models. J Am Stat Assoc 87(418):501–509CrossRefGoogle Scholar
  20. 20.
    Ferguson J (1980) Variable duration models for speech. In: Proceedings of the symposium on the application of hidden Markov models to text and speech, vol 1, pp 143–179Google Scholar
  21. 21.
    Fitzhugh R (1961) Impulses and physiological states in theoretical models of nerve membrane. Biophys J 1(6):445–466PubMedCentralPubMedCrossRefGoogle Scholar
  22. 22.
    Forney G Jr (1973) The viterbi algorithm. Proc IEEE 61(3):268–278CrossRefGoogle Scholar
  23. 23.
    Freudenthal A, Van Stuijvenberg M, Van Goudoever JB (2013) A quiet NICU for improved infants’ health, development and well-being: a systems approach to reducing noise and auditory alarms. Cogn Technol Work 15(3):329–345Google Scholar
  24. 24.
    Garcia SP, Almeida JS (2006) Multivariate phase space reconstruction by nearest neighbor embedding with different time delays. Phys Rev E 72(2):027205. doi:10.1103/PhysRevE.72.027205
  25. 25.
    Houtekamer P, Mitchell H (2001) A sequential ensemble Kalman filter for atmospheric data assimilation. Mon Weather Rev 129(2):123–137CrossRefGoogle Scholar
  26. 26.
    Imhoff M, Fried R (2009) The crying wolf: still crying? Anesth Analg 108(5):1382–1383CrossRefGoogle Scholar
  27. 27.
    Janvier A, Khairy M, Kokkotis A, Cormier C, Messmer D, Barrington K (2004) Apnea is associated with neurodevelopmental impairment in very low birth weight infants. J Perinatol 24(12):763–768PubMedCrossRefGoogle Scholar
  28. 28.
    Kalman R et al (1960) A new approach to linear filtering and prediction problems. J Basic Eng 82(1):35–45CrossRefGoogle Scholar
  29. 29.
    Kong L, Yang C, Wang Y, Gui W (2009) Generic phase space reconstruction method of multivariate time series. In: Control and decision conference, 2009. CCDC’09. Chinese, pp 3752–3755Google Scholar
  30. 30.
    Lawless ST (1994) Crying wolf: false alarms in a pediatric intensive care unit. Crit Care Med 22(6):981–985PubMedCrossRefGoogle Scholar
  31. 31.
    Lee HJ, Choi JH, Min SJ, Kim DH, Kim HS (2010) Comparison of the clinical performance between two pulse oximeters in nicu: Nellcor n-595 versus masimo set. J Korean Soc Neonatol 17(2):245–249CrossRefGoogle Scholar
  32. 32.
    Murphy K (2002) Dynamic bayesian networks: representation, inference and learning. Ph.D. thesis, CiteseerGoogle Scholar
  33. 33.
    Pichardo R, Adam JS, Rosow E, Bronzino J, Eisenfeld L (2003) Vibrotactile stimulation system to treat apnea of prematurity. Biomed Instrum Technol 37(1):34–40PubMedCrossRefGoogle Scholar
  34. 34.
    Pieczynski W (2003) Pairwise markov chains. IEEE Trans Pattern Anal Mach Intell 25(5):634–639. doi:10.1109/TPAMI.2003.1195998
  35. 35.
    Pillekamp F, Hermann C, Keller T, von Gontard A, Kribs A, Roth B (2007) Factors influencing apnea and bradycardia of prematurity-implications for neurodevelopment. Neonatology 91(3):155PubMedCrossRefGoogle Scholar
  36. 36.
    Poets C, Stebbens V, Samuels M, Southall D (1993) The relationship between bradycardia, apnea, and hypoxemia in preterm infants. Pediatr Res 34(2):144PubMedCrossRefGoogle Scholar
  37. 37.
    Portet F, Gao F, Hunter J, Sripada S (2007) Evaluation of on-line bradycardia boundary detectors from neonatal clinical data. In: Engineering in medicine and biology society, 2007. EMBS 2007. 29th annual international conference of the IEEE, pp 3288–3291Google Scholar
  38. 38.
    Rabiner LR (1989) A tutorial on hidden Markov models and selected applications in speech recognition. Proc IEEE 77(2):257–286CrossRefGoogle Scholar
  39. 39.
    Raman H, Sunilkumar N (1995) Multivariate modelling of water resources time series using artificial neural networks. Hydrol Sci J 40(2):145–164CrossRefGoogle Scholar
  40. 40.
    Schwarz G (1978) Estimating the dimension of a model. Ann Stat 6(2):461–464CrossRefGoogle Scholar
  41. 41.
    Suzuki T, Ikeguchi T, Suzuki M (2003) Multivariable nonlinear analysis of foreign exchange rates. Phys A Stat Mech Appl 323:591–600CrossRefGoogle Scholar
  42. 42.
    Task Force (1996) Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task force of the European society of cardiology and the north American society of pacing and electrophysiology. Circulation 93(5):1043–65Google Scholar
  43. 43.
    Viterbi A (1967) Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. IEEE Trans Inf Theory 13(2):260–269CrossRefGoogle Scholar
  44. 44.
    Yu SZ (2010) Hidden semi-Markov models. Artif Intell 174(2):215–243CrossRefGoogle Scholar
  45. 45.
    Zweig G, Russell S (1998) Speech recognition with dynamic Bayesian networks. In: Proceedings of the national conference on artificial intelligence, pp 173–180Google Scholar

Copyright information

© International Federation for Medical and Biological Engineering 2014

Authors and Affiliations

  • Miguel Altuve
    • 1
    • 2
  • Guy Carrault
    • 3
    • 4
  • Alain Beuchée
    • 3
    • 4
    • 5
  • Patrick Pladys
    • 3
    • 4
    • 5
  • Alfredo I. Hernández
    • 3
    • 4
  1. 1.Grupo de Bioingeniería y Biofísica AplicadaUniversidad Simón BolívarCaracasVenezuela
  2. 2.Departamento de Tecnología IndustrialUniversidad Simón Bolívar, Camurí GrandeCaracasVenezuela
  3. 3.Université de Rennes 1, LTSIRennesFrance
  4. 4.INSERM, U1099RennesFrance
  5. 5.Pôle Médico-Chirurgical de Pédiatrie et de Génétique Clinique, NéonatologieCHU RennesRennesFrance

Personalised recommendations