Use of genetic algorithms for neural networks to predict community-acquired pneumonia

BACKGROUND Genetic algorithms have been used to solve optimization problems for artificial neural networks (ANN) in several domains. We used genetic algorithms to search for optimal hidden-layer architectures, connectivity, and training parameters for ANN for predicting community-acquired pneumonia among patients with respiratory complaints. METHODS Feed-forward back-propagation ANN were trained on sociodemographic, symptom, sign, comorbidity, and radiographic outcome data among 1044 patients from the University of Illinois (the training cohort), and were applied to 116 patients from the University of Nebraska (the testing cohort). Binary chromosomes with genes representing network attributes, including the number of nodes in the hidden layers, learning rate and momentum parameters, and the presence or absence of implicit within-layer connectivity using a competition algorithm, were operated on by various combinations of crossover, mutation, and probabilistic selection based on network mean-square error (MSE), and separately on average cross entropy (ENT). Predictive accuracy was measured as the area under a receiver-operating characteristic (ROC) curve. RESULTS Over 50 generations, the baseline genetic algorithm evolved an optimized ANN with nine nodes in the first hidden layer, zero nodes in the second hidden layer, learning rate and momentum parameters of 0.5, and no within-layer competition connectivity. This ANN had an ROC area in the training cohort of 0.872 and in the testing cohort of 0.934 (P-value for difference, 0.181). Algorithms based on cross-generational selection, Gray coding of genes prior to mutation, and crossover recombination at different genetic levels, evolved optimized ANN identical to the baseline genetic strategy. Algorithms based on other strategies, including elite selection within generations (training ROC area 0.819), and inversions of genetic material during recombination (training ROC area 0.812), evolved less accurate ANN. CONCLUSION ANN optimized by genetic algorithms accurately discriminated pneumonia within a training cohort, and within a testing cohort consisting of cases on which the networks had not been trained. Genetic algorithms can be used to implement efficient search strategies for optimal ANN to predict pneumonia.

[1]  A. El‐Solh,et al.  Predicting active pulmonary tuberculosis using an artificial neural network. , 1999, Chest.

[2]  W. Baxt.  Use of an artificial neural network for the diagnosis of myocardial infarction. , 1991, Annals of internal medicine.

[3]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[4]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[5]  R. Dybowski,et al.  Prediction of outcome in critically ill patients using artificial neural network synthesised by genetic algorithm , 1996, The Lancet.

[6]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[7]  J. Hanley,et al.  A method of comparing the areas under receiver operating characteristic curves derived from the same cases. , 1983, Radiology.

[8]  J P Ornato,et al.  Use of Clinical Judgment Analysis to Explain Regional Variations in Physicians' Accuracies in Diagnosing Pneumonia , 1991, Medical decision making : an international journal of the Society for Medical Decision Making.

[9]  M. Rubenfire,et al.  Neural network in the clinical diagnosis of acute pulmonary embolism. , 1993, Chest.

[10]  P. Wilding,et al.  The application of backpropagation neural networks to problems in pathology and laboratory medicine. , 1992, Archives of pathology & laboratory medicine.

[11]  N Pendleton,et al.  New approach to risk determination: development of risk profile for new falls among community-dwelling older people by use of a Genetic Algorithm Neural Network (GANN). , 2000, The journals of gerontology. Series A, Biological sciences and medical sciences.

[12]  Sungzoon Cho,et al.  Multiple disorder diagnosis with adaptive competitive neural networks , 1993, Artif. Intell. Medicine.

[13]  L. Tucker.  A SUGGESTED ALTERNATIVE FORMULATION IN THE DEVELOPMENTS BY HURSCH, HAMMOND, AND HURSCH, AND BY HAMMOND, HURSCH, AND TODD. , 1964, Psychological review.

[14]  Thomas Roß,et al.  Feature selection for optimized skin tumor recognition using genetic algorithms , 1999, Artif. Intell. Medicine.

[15]  R. Wigton,et al.  Prediction of Community-Acquired Pneumonia Using Artificial Neural Networks , 2003, Medical decision making : an international journal of the Society for Medical Decision Making.

[16]  C. D. Gelatt,et al.  Optimization by Simulated Annealing , 1983, Science.

[17]  S Forrest,et al.  Genetic algorithms , 1996, CSUR.

[18]  James L. McClelland.  Explorations In Parallel Distributed Processing , 1988 .

[19]  K. R. Hammond,et al.  ANALYZING THE COMPONENTS OF CLINICAL INFERENCE. , 1964, Psychological review.

[20]  J V Tu,et al.  Advantages and disadvantages of using artificial neural networks versus logistic regression for predicting medical outcomes. , 1996, Journal of clinical epidemiology.

[21]  Paul S. Heckerling,et al.  Parametric receiver operating characteristic curve analysis using mathematica , 2002, Comput. Methods Programs Biomed..

[22]  Lawrence S. Kroll.  Mathematica--A System for Doing Mathematics by Computer. , 1989 .

[23]  W Penny,et al.  Neural Networks in Clinical Medicine , 1996, Medical decision making : an international journal of the Society for Medical Decision Making.

[24]  G. Kane.  Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol 1: Foundations, vol 2: Psychological and Biological Models , 1994 .

[25]  A Gottschalk,et al.  A Comparison of Human and Machine-based Predictions of Successful Weaning from Mechanical Ventilation , 2000, Medical decision making : an international journal of the Society for Medical Decision Making.

[26]  Paulo J. G. Lisboa,et al.  A review of evidence of health benefit from artificial neural networks in medical intervention , 2002, Neural Networks.

[27]  Laurene V. Fausett,et al.  Fundamentals Of Neural Networks , 1994 .

[28]  J P Ornato,et al.  Clinical prediction rule for pulmonary infiltrates. , 1990, Annals of internal medicine.

[29]  M F Jefferson,et al.  Comparison of a genetic algorithm neural network with logistic regression for predicting outcome after surgery for patients with nonsmall cell lung carcinoma , 1997, Cancer.

[30]  W. Grove.  Statistical Methods for Rates and Proportions, 2nd ed , 1981 .

[31]  M F Jefferson,et al.  Evolution of Artificial Neural Network Architecture: Prediction of Depression after Mania , 1998, Methods of Information in Medicine.

[32]  C. Floyd,et al.  Acute pulmonary embolism: artificial neural network approach for diagnosis. , 1993, Radiology.

[33]  M. Narayanan,et al.  A Genetic Algorithm to Improve a Neural Network to Predict a Patient’s Response to Warfarin , 1993, Methods of Information in Medicine.

[34]  Stephen Wolfram,et al.  Mathematica: a system for doing mathematics by computer (2nd ed.) , 1991 .

[35]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[36]  D. Dorfman,et al.  Maximum-likelihood estimation of parameters of signal-detection theory and determination of confidence intervals—Rating-method data , 1969 .

[37]  Michael J Fine,et al.  Testing Strategies in the Initial Management of Patients with Community-Acquired Pneumonia , 2002, Annals of Internal Medicine.

[38]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[39]  E. Brunswik.  Perception and the Representative Design of Psychological Experiments , 1957 .

[40]  B. Everitt,et al.  Statistical methods for rates and proportions , 1973 .

[41]  James A. Freeman,et al.  Simulating neural networks - with Mathematica , 1993 .

[42]  Bruce W. Colletti,et al.  Artificial intelligence versus logistic regression statistical modelling to predict cardiac complications after noncardiac surgery , 1994, Clinical Cardiology.

[43]  R W Veltri,et al.  Genetically engineered neural networks for predicting prostate cancer progression after radical prostatectomy. , 1999, Urology.

[44]  Christian Jacob,et al.  Illustrating Evolutionary Computation with Mathematica , 2001 .