Skip to main content
PMC home page
Archives of Disease in Childhood. Fetal and Neonatal Edition logoLink to Archives of Disease in Childhood. Fetal and Neonatal Edition
. 1998 Sep;79(2):F129–F134. doi: 10.1136/fn.79.2.f129

Artificial neural network for risk assessment in preterm neonates

B Zernikow, K Holtmannspoetter, E Michel, W Pielemeier, F Hornschuh, A Westermann, K Hennecke
PMCID: PMC1720838  PMID: 9828740

Abstract

AIM—To predict the individual neonatal mortality risk of preterm infants using an artificial neural network "trained" on admission data.
METHODS—A total of 890 preterm neonates (<32 weeks gestational age and/or <1500 g birthweight) were enrolled in our retrospective study. The neural network trained on infants born between 1990and 1993. The predictive value was tested on infants born in the successive three years.
RESULTS—The artificial neural network performed significantly better than a logistic regression model (area under the receiver operator curve 0.95 vs 0.92). Survival was associated with high morbidity if the predicted mortality risk was greater than 0.50. There were no preterm infants with a predicted mortality risk of greater than 0.80. The mortality risks of two non-survivors with birthweights >2000 g and severe congenital disease had largely been underestimated.
CONCLUSION—An artificial neural network trained on admission data can accurately predict the mortality risk for most preterm infants. However, the significant number of prediction failures renders it unsuitable for individual treatment decisions.



Full Text

The Full Text of this article is available as a PDF (130.8 KB).

Selected References

These references are in PubMed. This may not be the complete list of references from this article.

  1. Baxt W. G. Application of artificial neural networks to clinical medicine. Lancet. 1995 Oct 28;346(8983):1135–1138. doi: 10.1016/s0140-6736(95)91804-3. [DOI] [PubMed] [Google Scholar]
  2. Baxt W. G., Skora J. Prospective validation of artificial neural network trained to identify acute myocardial infarction. Lancet. 1996 Jan 6;347(8993):12–15. doi: 10.1016/s0140-6736(96)91555-x. [DOI] [PubMed] [Google Scholar]
  3. Boyd O., Grounds R. M. Physiological scoring systems and audit. Lancet. 1993 Jun 19;341(8860):1573–1574. doi: 10.1016/0140-6736(93)90706-m. [DOI] [PubMed] [Google Scholar]
  4. Cooke R. W. Trends in preterm survival and incidence of cerebral haemorrhage 1980-9. Arch Dis Child. 1991 Apr;66(4 Spec No):403–407. doi: 10.1136/adc.66.4_spec_no.403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Corbet A., Gerdes J., Long W., Avila E., Puri A., Rosenberg A., Edwards K., Cook L. Double-blind, randomized trial of one versus three prophylactic doses of synthetic surfactant in 826 neonates weighing 700 to 1100 grams: effects on mortality rate. American Exosurf Neonatal Study Groups I and IIa. J Pediatr. 1995 Jun;126(6):969–978. doi: 10.1016/s0022-3476(95)70226-1. [DOI] [PubMed] [Google Scholar]
  6. Cross S. S., Harrison R. F., Kennedy R. L. Introduction to neural networks. Lancet. 1995 Oct 21;346(8982):1075–1079. doi: 10.1016/s0140-6736(95)91746-2. [DOI] [PubMed] [Google Scholar]
  7. Dybowski R., Gant V. Artificial neural networks in pathology and medical laboratories. Lancet. 1995 Nov 4;346(8984):1203–1207. doi: 10.1016/s0140-6736(95)92904-5. [DOI] [PubMed] [Google Scholar]
  8. Dybowski R., Weller P., Chang R., Gant V. Prediction of outcome in critically ill patients using artificial neural network synthesised by genetic algorithm. Lancet. 1996 Apr 27;347(9009):1146–1150. doi: 10.1016/s0140-6736(96)90609-1. [DOI] [PubMed] [Google Scholar]
  9. Ebell M. H. Artificial neural networks for predicting failure to survive following in-hospital cardiopulmonary resuscitation. J Fam Pract. 1993 Mar;36(3):297–303. [PubMed] [Google Scholar]
  10. Forsström J. J., Dalton K. J. Artificial neural networks for decision support in clinical medicine. Ann Med. 1995 Oct;27(5):509–517. doi: 10.3109/07853899509002462. [DOI] [PubMed] [Google Scholar]
  11. Gerstmann D. R., Minton S. D., Stoddard R. A., Meredith K. S., Monaco F., Bertrand J. M., Battisti O., Langhendries J. P., Francois A., Clark R. H. The Provo multicenter early high-frequency oscillatory ventilation trial: improved pulmonary and clinical outcome in respiratory distress syndrome. Pediatrics. 1996 Dec;98(6 Pt 1):1044–1057. [PubMed] [Google Scholar]
  12. Hanley J. A., McNeil B. J. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982 Apr;143(1):29–36. doi: 10.1148/radiology.143.1.7063747. [DOI] [PubMed] [Google Scholar]
  13. Horbar J. D., McAuliffe T. L., Adler S. M., Albersheim S., Cassady G., Edwards W., Jones R., Kattwinkel J., Kraybill E. N., Krishnan V. Variability in 28-day outcomes for very low birth weight infants: an analysis of 11 neonatal intensive care units. Pediatrics. 1988 Oct;82(4):554–559. [PubMed] [Google Scholar]
  14. Horbar J. D., Onstad L., Wright E. Predicting mortality risk for infants weighing 501 to 1500 grams at birth: a National Institutes of Health Neonatal Research Network report. Crit Care Med. 1993 Jan;21(1):12–18. doi: 10.1097/00003246-199301000-00008. [DOI] [PubMed] [Google Scholar]
  15. Maier R. F., Rey M., Metze B. C., Obladen M. Comparison of mortality risk: a score for very low birthweight infants. Arch Dis Child Fetal Neonatal Ed. 1997 May;76(3):F146–F151. doi: 10.1136/fn.76.3.f146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Metz C. E. Basic principles of ROC analysis. Semin Nucl Med. 1978 Oct;8(4):283–298. doi: 10.1016/s0001-2998(78)80014-2. [DOI] [PubMed] [Google Scholar]
  17. Patterson C. C., Halliday H. L. Prediction of outcome shortly after delivery for the very low birthweight (less than or equal to 1500 g) infant. Paediatr Perinat Epidemiol. 1988 Jul;2(3):221–228. doi: 10.1111/j.1365-3016.1988.tb00211.x. [DOI] [PubMed] [Google Scholar]
  18. Perlman M., Claris O., Hao Y., Pandit P., Whyte H., Chipman M., Liu P. Secular changes in the outcomes to eighteen to twenty-four months of age of extremely low birth weight infants, with adjustment for changes in risk factors and severity of illness. J Pediatr. 1995 Jan;126(1):75–87. doi: 10.1016/s0022-3476(95)70507-4. [DOI] [PubMed] [Google Scholar]
  19. Phelps D. L., Brown D. R., Tung B., Cassady G., McClead R. E., Purohit D. M., Palmer E. A. 28-day survival rates of 6676 neonates with birth weights of 1250 grams or less. Pediatrics. 1991 Jan;87(1):7–17. [PubMed] [Google Scholar]
  20. Rautonen J., Mäkelä A., Boyd H., Apajasalo M., Pohjavuori M. CRIB and SNAP: assessing the risk of death for preterm neonates. Lancet. 1994 May 21;343(8908):1272–1273. doi: 10.1016/s0140-6736(94)92158-x. [DOI] [PubMed] [Google Scholar]
  21. Richardson D. K., Gray J. E., McCormick M. C., Workman K., Goldmann D. A. Score for Neonatal Acute Physiology: a physiologic severity index for neonatal intensive care. Pediatrics. 1993 Mar;91(3):617–623. [PubMed] [Google Scholar]
  22. Richardson D. K., Phibbs C. S., Gray J. E., McCormick M. C., Workman-Daniels K., Goldmann D. A. Birth weight and illness severity: independent predictors of neonatal mortality. Pediatrics. 1993 May;91(5):969–975. [PubMed] [Google Scholar]
  23. Schmidt B., Kirpalani H., Rosenbaum P., Cadman D. Strengths and limitations of the Apgar score: a critical appraisal. J Clin Epidemiol. 1988;41(9):843–850. doi: 10.1016/0895-4356(88)90100-x. [DOI] [PubMed] [Google Scholar]
  24. Stevens S. M., Richardson D. K., Gray J. E., Goldmann D. A., McCormick M. C. Estimating neonatal mortality risk: an analysis of clinicians' judgments. Pediatrics. 1994 Jun;93(6 Pt 1):945–950. [PubMed] [Google Scholar]
  25. Tarnow-Mordi W., Ogston S., Wilkinson A. R., Reid E., Gregory J., Saeed M., Wilkie R. Predicting death from initial disease severity in very low birthweight infants: a method for comparing the performance of neonatal units. BMJ. 1990 Jun 23;300(6740):1611–1614. doi: 10.1136/bmj.300.6740.1611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Verloove-Vanhorick S. P., Verwey R. A., Brand R., Gravenhorst J. B., Keirse M. J., Ruys J. H. Neonatal mortality risk in relation to gestational age and birthweight. Results of a national survey of preterm and very-low-birthweight infants in the Netherlands. Lancet. 1986 Jan 11;1(8472):55–57. doi: 10.1016/s0140-6736(86)90713-0. [DOI] [PubMed] [Google Scholar]

Articles from Archives of Disease in Childhood. Fetal and Neonatal Edition are provided here courtesy of BMJ Publishing Group

RESOURCES