Backpropagation artificial neural network to detect hyperthermic seizures in rats
Sinha, Rakesh Kumar (2003) Backpropagation artificial neural network to detect hyperthermic seizures in rats. Online Journal of Health and Allied Sciences, 1 (4). ISSN 0972-5997
Full text available as:
A three-layered feed-forward back-propagation Artificial Neural Network was used to classify the seizure episodes in rats. Seizure patterns were induced by subjecting anesthetized rats to a Biological Oxygen Demand incubator at 45-47ºC for 30 to 60 minutes. Selected fast Fourier transform data of one second epochs of electroencephalogram were used to train and test the network for the classification of seizure and normal patterns. The results indicate that the present network with the architecture of 40-12-1 (input-hidden-output nodes) agrees with manual scoring of seizure and normal patterns with a high recognition rate of 98.6%.
Archive Staff Only: edit this record