The aim of this research was to apply a generalized regression neural network (GRNN) to predict neutron spectrum using the rates count coming from a Bonner spheres system as the only piece of information. In the training and testing stages, a data set of 251 different types of neutron spectra, taken from the International Atomic Energy Agency compilation, were used. Fifty-one predicted spectra were analyzed at testing stage. Training and testing of GRNN were carried out in the MATLAB environment by means of a scientific and technological tool designed based on GRNN technology, which is capable of solving the neutron spectrometry problem with high performance and generalization capability. This computational tool automates the pre-processing of information, the training and testing stages, the statistical analysis, and the post-processing of the information. In this work, the performance of feed-forward backpropagation neural networks (FFBPNN) and GRNN was compared in the solution of the neutron spectrometry problem. From the results obtained, it can be observed that despite very similar results, GRNN performs better than FFBPNN because the former could be used as an alternative procedure in neutron spectrum unfolding methodologies with high performance and accuracy.
Part of the book: Artificial Neural Networks
Breast cancer is one of the global leading causes of death among women, and an early detection is of uttermost importance to reduce mortality rates. Screening mammograms, in which radiologists rely only on their eyesight, are one of the most used early detection methods. However, characteristics, such as the asymmetry between breasts, a feature that could be very difficult to visually quantize, is key to breast cancer detection. Due to the highly heterogeneous and deformable structure of the breast itself, incorporating asymmetry measurements into an automated detection system is still a challenge. In this study, we proposed the use of a bilateral registration algorithm as an effective way to automatically measure mirror asymmetry. Furthermore, this information was fed to a machine learning algorithm to improve the accuracy of the model. In this study, 449 subjects (197 with calcifications, 207 with masses, and 45 healthy subjects) from a public database were used to train and evaluate the proposed methodology. Using this procedure, we were able to independently identify subjects with calcifications (accuracy = 0.825, AUC = 0.882) and masses (accuracy = 0.698, AUC = 0.807) from healthy subjects.
Part of the book: Breast Imaging