BACKGROUND: Artificial neural networks (ANN) are gaining prominence as a method of classification in a wide range of disciplines. In this study ANN is applied to data from a verbal autopsy study as a means of classifying cause of death. METHODS: A simulated ANN was trained on a subset of verbal autopsy data, and the performance was tested on the remaining data. The performance of the ANN models were compared to two other classification methods (physician review and logistic regression) which have been tested on the same verbal autopsy data. RESULTS: Artificial neural network models were as accurate as or better than the other techniques in estimating the cause-specific mortality fraction (CSMF). They estimated the CSMF within 10% of true value in 8 out of 16 causes of death. Their sensitivity and specificity compared favourably with that of data-derived algorithms based on logistic regression models. CONCLUSIONS: Cross-validation is crucial in preventing the over-fitting of the ANN models to the training data. Artificial neural network models are a potentially useful technique for classifying causes of death from verbal autopsies. Large training data sets are needed to improve the performance of data-derived algorithms, in particular ANN models.
BACKGROUND: Artificial neural networks (ANN) are gaining prominence as a method of classification in a wide range of disciplines. In this study ANN is applied to data from a verbal autopsy study as a means of classifying cause of death. METHODS: A simulated ANN was trained on a subset of verbal autopsy data, and the performance was tested on the remaining data. The performance of the ANN models were compared to two other classification methods (physician review and logistic regression) which have been tested on the same verbal autopsy data. RESULTS: Artificial neural network models were as accurate as or better than the other techniques in estimating the cause-specific mortality fraction (CSMF). They estimated the CSMF within 10% of true value in 8 out of 16 causes of death. Their sensitivity and specificity compared favourably with that of data-derived algorithms based on logistic regression models. CONCLUSIONS: Cross-validation is crucial in preventing the over-fitting of the ANN models to the training data. Artificial neural network models are a potentially useful technique for classifying causes of death from verbal autopsies. Large training data sets are needed to improve the performance of data-derived algorithms, in particular ANN models.
Authors: Ana L Garces; Elizabeth M McClure; Wilton Pérez; K Michael Hambidge; Nancy F Krebs; Lester Figueroa; Carl L Bose; Waldemar A Carlo; Constance Tenge; Fabian Esamai; Shivaprasad S Goudar; Sarah Saleem; Archana B Patel; Melody Chiwila; Elwyn Chomba; Antoinette Tshefu; Richard J Derman; Patricia L Hibberd; Sherri Bucher; Edward A Liechty; Melissa Bauserman; Janet L Moore; Marion Koso-Thomas; Menachem Miodovnik; Robert L Goldenberg Journal: Acta Paediatr Date: 2017-04-05 Impact factor: 2.299
Authors: Rafael Lozano; Michael K Freeman; Spencer L James; Benjamin Campbell; Alan D Lopez; Abraham D Flaxman; Christopher Jl Murray Journal: Popul Health Metr Date: 2011-08-05
Authors: Philip W Setel; Chalapati Rao; Yusuf Hemed; David R Whiting; Gonghuan Yang; Daniel Chandramohan; K G M M Alberti; Alan D Lopez Journal: PLoS Med Date: 2006-08 Impact factor: 11.069