R. Dybowski et al., PREDICTION OF OUTCOME IN CRITICALLY ILL PATIENTS USING ARTIFICIAL NEURAL-NETWORK SYNTHESIZED BY GENETIC ALGORITHM, Lancet, 347(9009), 1996, pp. 1146-1150
Background Decisions about which patients to admit to intensive care a
nd how long to keep them there are difficult. A flexible computer-base
d mathematical model which is sensitive to the complexity of intensive
care medicine, and which accurately models prognosis, seems highly de
sirable. Methods We have created, optimised by genetic algorithms, tra
ined, and evaluated the performance of an artificial neural network (A
NN) in the clinical setting of systemic inflammatory response syndrome
and haemodynamic shock. 258 patients were selected from an intensive
care database of 4484 patients al a London teaching hospital and rando
mised to a network training set (168) and a test set (90). The outcome
evaluated was death during that hospital admission and the performanc
e of the neural net was compared (by receiver operating characteristic
[ROC] curves and by Brier scores) with that of a logistic regression
model. Findings Artificial neural network performance increased with s
uccessive generations; the best-performing ANN was created after 7 gen
erations and predicted outcome more accurately than the logistic regre
ssion model (ROC curve area 0.863 vs 0.753). Interpretation In this st
udy, ANNs have lent themselves particularly well to modelling a comple
x clinical situation; we suggest that this relates to their inherently
flexible nature which accommodates interactions between the clinical
input fields. In addition, we have demonstrated the value of a second
computational technique (genetic algorithms) in ''tuning'' ANN perform
ance. These techniques can potentially be implemented in individual in
tensive care units; the outcome models which they will generate will b
e sensitive to local practice. Analysis of such accurate clinical outc
ome models may empower clinicians with a hitherto unappreciated degree
of insight into those elements of their clinical practice which are m
ost relevant to their patients' outcome.