Diagnostic classification of undifferentiated fevers using artificial neural network

Accurate diagnosis of undifferentiated fever case at the earliest is a challenging effort, which needs extensive diagnostic tests. Prediction of undifferentiated fever cases at an early stage will help in diagnosing the disease in comparatively lesser time and more effectively. The aim of the present study was to apply Artificial Intelligence (AI) algorithm using temperature information for the prediction of major categories of diseases among undifferentiated fever cases. This was an observational study carried out in tertiary care hospital. Total of 103 patients were involved in the study and 24-hour continuous temperature recording was done. Analysis was done using Artificial Neural Network (ANN) model based on the temperature data of each patients and its statistical parameters. Temperature datasets were labeled with the help of experienced physicians. Levenberg Marquardt error back-propagation algorithm was used to train the network. A good relation was found between the target data set and output data set, purely based on the observed 24 hr continuous tympanic temperature of the patients. An accuracy of 98.1% was obtained from ANN prediction model. The study concluded that a single noninvasive temperature parameter is sufficient to predict the major categories of diseases using ANN algorithms, from the undifferentiated fever cases.