Abstract
The subjectiveness of patient data and the incompleteness of doctors’ causal models has led to a wide variety of techniques being used to build medical decision-aids. We briefly describe problems resulting from the use of conventional methods for building an advisor for chest pain, before exploring the way in which doctors’ decisions might be modelled using connectionist (neural network) techniques. We then examine the claims and the realities of current connectionist models. To determine if such techniques are promising for this domain, a series of proven 3-layer and a 4-layer backward- propagation networks were repeatedly trained on data from 174 chest pain cases. The best networks were then assessed on fresh data from a new set of 73 cases. The 3-layer network performed better, but still had a clinically unacceptable crude accuracy of 70% and false negative rate of 27%. We are concerned about the undisciplined use of connectionist systems, and about the problems of system validation. There is a need for further work on the provision of informative diagnostics and statistics for data exploration and training. Given this greater understanding, connectionist systems might have much to offer as components of medical decision-aids.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Bounds DG, Lloyd PJ, Mathew B, Waddell G, 1988. A multi layer perceptron network for the diagnosis of low back pain. Proc. San Diego conference on neural networks, vol. 2, pp. 481–489.
Breiman L, Friedman JH, Olshen RA and Stone CJ, 1984. Classification and Regression Trees. Wadsworth International: Belmont California.
Deleu J, Beuscart R, Becquast E, Duhamel A, Comyn G, 1988. Reseu Bayesien et diagnostique medicale. Proc. Int. workshop on neural networks and applications, Nimes, November 1988; pp. 95–103
Emerson PA, Russell N, Wyatt J, Crichton NJ. 1989. An audit of the management of patients attending an accident & emergency department with chest pain; QJ Med. (in press)
Emerson PA, Wyatt J, Dillistone L, Crichton N, Russell N, 1988. The development of ACORN, an expert system to enable nurses to make decisions about casualty patients with chest pain; in proc. Medical Informatics: computers in clinical medicne, pub. British Medical Informatics Soc, London.
Gallant SI 1988. Connectionist Expert Systems. Comm. ACM vol 31, no 2 pp 152–166
Hopfield JJ, Tank 1986. Computing with neural circuits: a model. Science 1986; 233: 625–633.
Kolmogorov, AN 1956. On the representation of continuous functions of several variables by superposition of continuous functions of a smaller number of variables. Dokl. Akad. SSSR 108 179–182
Laszlo G 1977. Investigation of pulmonary function: current practice in 10 provincial British laboratories; Proc.Roy.Soc. Med. vol. 70, pp. 163–5.
Lauritzen SL, Spiegelhalter DJ 1988. Local computation on graphical structures and their application to expert systems. J RSS series B, 50:157–224
Lippman RP 1987. An introduction to computing with neural nets. IEEE ASSp Mag. April 1987; pp. 4–22
McClelland JL Rumelhart DE & the PDP research group, 1986. Parallel distributed processing (vols 1 and 2). MIT Press, Boston
Manly BFJ 1986. Multivariate Statistical Methods: A Primer. Chapman and Hall, London.
Saunders K 1977. Current practice in six London lung function laboratories; Proc.Roy.Soc. Med. vol. 70, pp. 162–3
Spiteri MA, Cook DG, Clarke SW 1988. Reliability of eliciting physical signs in examination of the chest; Lancet 16/4/88, pp 873–875.
Sox HC, Blatt MA, Higgins MC, Marton KI, 1988. Medical decision making; pub. Butterworths, Boston.
Szolovits P & Pauker S, 1978. Categorical and probabilistic reasoning in medical diagnosis; Artificial Intelligence vol. 11, pp. 115–144
Wyatt J, Emerson P, 1989. A pragmatic approach to knowledge engineering, with examples of use in a difficult domain; in Expert Systems — Human Issues, eds. Hart A & Berry D, pub. Kogan Page
Wyatt J 1988. Thrombolysis in acute MI: why the delay ?; Brit. Heart J vol. 59, p 618.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1989 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Hart, A., Wyatt, J. (1989). Connectionist models in medicine: an investigation of their potential. In: Hunter, J., Cookson, J., Wyatt, J. (eds) AIME 89. Lecture Notes in Medical Informatics, vol 38. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-93437-7_15
Download citation
DOI: https://doi.org/10.1007/978-3-642-93437-7_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-51543-2
Online ISBN: 978-3-642-93437-7
eBook Packages: Springer Book Archive