Skip to main content
Log in

Künstliche neuronale Netze

Theorie und Anwendungen in der Anästhesie, Intensiv- und Notfallmedizin

  • Trends und Medizinökonomie
  • Published:
Der Anaesthesist Aims and scope Submit manuscript

Zusammenfassung

Mit einem künstlichen neuronalen Netzwerk (KNN) wird versucht, die Vorgänge im Zentralnervensystem (ZNS) höherer Lebewesen zu simulieren. In aller Regel erfolgt dies durch eine spezielle Software, die das Verhalten einzelner Neurone und deren Interaktion miteinander simuliert. Der wesentliche Unterschied zu klassischen statistischen Berechnungsverfahren besteht in der Lernfähigkeit eines KNN. Das bedeutet, dass ein solches System anfänglich keinerlei Informationen enthält, sondern sich diese erst aus einer gewissen Zahl bekannter Beispiele extrahiert. Im Idealfall kann ein KNN durch wiederholtes Training generalisieren, d. h., es verändert die Verknüpfungen innerhalb des Neuronenverbandes so, dass es später unbekannte Daten anhand der erlernten Regeln richtig klassifizieren kann. Darüber hinaus reagieren neuronale Netze wenig empfindlich gegen gestörte oder unvollständige Daten. Künstliche neuronale Netze haben bereits in anderen Bereichen gezeigt, dass sie zur Vorhersage von Ereignissen und Modellierung komplexer zeitabhängiger Systeme geeignet sind. Außerhalb der Medizin werden sie v. a. dann eingesetzt, wenn die Einflussfaktoren für einen bestimmten Ausgang nicht oder nur unvollständig bekannt und die Zusammenhänge komplex und nichtlinear sind (z. B. bei Finanz- oder Wetterprognosen). Dieser Artikel soll eine kurze Übersicht über die grundsätzliche Funktionsweise von KNN geben und potenzielle Einsatzmöglichkeiten in Anästhesie, Intensiv- und Notfallmedizin aufzeigen.

Summary

Artificial neural networks (ANN) are constructed to simulate processes of the central nervous system of higher creatures. An ANN consists of a set of processing units (nodes) which simulate neurons and are interconnected via a set of "weights" (analogous to synaptic connections in the nervous system) in a way which allows signals to travel through the network in parallel. The nodes (neurons) are simple computing elements. They accumulate input from other neurons by means of a weighted sum. If a certain threshold is reached the neuron sends information to all other connected neurons otherwise it remains quiescent. One major difference compared with traditional statistical or rule-based systems is the learning aptitude of an ANN. At the very beginning of a training process an ANN contains no explicit information. Then a large number of cases with a known outcome are presented to the system and the weights of the inter-neuronal connections are changed by a training algorithm designed to minimise the total error of the system. A trained network has extracted rules that are represented by the matrix of the weights between the neurons. This feature is called generalisation and allows the ANN to predict cases that have never been presented to the system before. Artificial neural networks have shown to be useful predicting various events. Especially complex, non-linear, and time depending relationships can be modelled and forecasted. Furthermore an ANN can be used when the influencing variables on a certain event are not exactly known as it is the case in financial or weather forecasts. This article aims to give a short overview on the function of ANN and their previous use and possible future applications in anaesthesia, intensive care, and emergency medicine.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Abb. 1
Abb. 2
Abb. 3

Literatur

  1. Baxt WG (1991) Use of an artificial neural network for the diagnosis of myocardial infarction. Ann Intern Med 115:843–848

    CAS  Google Scholar 

  2. Baxt WG (1992) Analysis of the clinical variables driving decision in an artificial neural network trained to identify the presence of myocardial infarction. Ann Emerg Med 21:1439–1444

    CAS  Google Scholar 

  3. Baxt WG, Shofer FS, Sites FD, Hollander JE (2002) A neural computational aid to the diagnosis of acute myocardial infarction. Ann Emerg Med 39:366–373

    Article  PubMed  Google Scholar 

  4. Dybowski R, Weller P, Chang R, Gant V (1996) Prediction of outcome in critically ill patients using artificial neural network synthesised by genetic algorithm. Lancet 347:1146–1150

    CAS  Google Scholar 

  5. Ebell H (1993) Artificial neural networks for predicting failure to survive following in-hospital cardiopulmonary resuscitation. J Fam Pract 36:297–303

    CAS  PubMed  Google Scholar 

  6. Flanagan JR, Pittet D, Li N, Thievent B, Suter PM, Wenzel RP (1996) Predicting survival of patients with sepsis by use of regression and neural network models. Perform Qual Health Care 4:96–103

    CAS  Google Scholar 

  7. Geldner G, Schwarz U, Ruoff M, Romeiser J, Lendl M, Schutz W, Georgieff M (1999) Entwicklung eines neuen Closed-Loop Systems. Regelung einer Mivacurium-indizierten neuromuskulären Blockade. Anaesthesist 48:157–162

    CAS  Google Scholar 

  8. Kamangar FA, Behbehani K (1997) An artificial neural network-based controller for the control of induced paralysis using vecuronium bromide. Ann Biomed Eng 25:1040–1051

    CAS  PubMed  Google Scholar 

  9. Kennedy RL, Harrison RF, Burton AM et al. (1997) An artificial neural network system for diagnosis of acute myocardial infarction (AMI) in the accident and emergency department: evaluation and comparison with serum myoglobin measurements. Comput Methods Programs Biomed 52:93–103

    Article  CAS  PubMed  Google Scholar 

  10. Laffey JG, Tobin E, Boylan JF, McShane AJ (2003) Assessment of a simple artificial neural network for predicting residual neuromuscular block. Br J Anaesth 90:48–52

    Article  CAS  PubMed  Google Scholar 

  11. Lawrence J (1992) Neuronale Netze—Computersimulation biologischer Intelligenz. Systhema, München

  12. Patterson, Dan W (1997) Künstliche neuronale Netze—Das Lehrbuch. Prentice Hall, München London Mexiko

  13. Rutledge R (1995) Injury severity and probability of survival assessment in trauma patients using a predictive hierarchical network model derived from ICD-9 codes. J Trauma 38:590–601

    CAS  PubMed  Google Scholar 

  14. Santanen OAP, Svartling N, Haasio J, Paloheimo MPJ (2003) Neural nets and prediction of the recovery rate from neuromuscular block. Eur J Anaesth 20:87–92

    CAS  Google Scholar 

  15. Tu JV (1996) Advantages and disadvantages of using artificial neural networks versus logistic regression for predicting medical outcomes. J Clin Epidemiol 49:1225–1231

    Article  CAS  PubMed  Google Scholar 

  16. Tu JV, Guerriere MRJ (1993) Use of a neural network as a predictive instrument for length of stay in the intensive care unit following cardiac surgery. Comput Biomed Res 26:220–229

    Article  CAS  PubMed  Google Scholar 

  17. Tu JV, Mazer CD, Levinton C, Armstrong PW, Naylor CD (1994) A predictive index for length of stay in the intensive care unit following cardiac surgery. Can Med Assoc J 151:177–185

    CAS  Google Scholar 

  18. Wong LS, Young JD (1999) A comparison of ICU mortality prediction using the APACHE II scoring system and artificial neural networks. Anaesthesia 54:1048–1054

    CAS  Google Scholar 

  19. Zell A (1996) Simulation neuronaler Netze. Addison-Wesley, München

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to L. H. J. Eberhart.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Traeger, M., Eberhart, A., Geldner, G. et al. Künstliche neuronale Netze. Anaesthesist 52, 1055–1061 (2003). https://doi.org/10.1007/s00101-003-0576-x

Download citation

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00101-003-0576-x

Schlüsselwörter

Keywords

Navigation