Minimize the mean square error by data segregation approach for back-propagation artificial neural network with adaptive learning based image reconstruction in electron magnetic resonance imaging tomography

This paper presents the data segregation strategies applied on a back-propagation artificial neural network (BP-ANN) with adaptive learning algorithm. The application system is developed for reconstruction of two-dimensional spatial images from continuous wave electron magnetic resonance imaging (CW-EMRI) tomography data. We propose that the exemplar datasets to be segregated into subsets. Using these subsets, artificial sub neural nets (subnets) are constructed and training is carried out. The proposed method yields better PSNR values and less mean square error values. The performance results are tabulated for different subnet sizes.