Last friday, the MSc student Caio Oliveira Carneloz defended his master’s thesis entitled “Auxílio no diagnóstico da doença de Alzheimer a partir de imagens de ressonância magnética utilizando competição e cooperação entre partículas” (“Diagnostic aid for Alzheimer’s disease from magnetic resonance imaging using particle competition and cooperation”), and he was approved. Congratulations, Caio!

In his work, he presents a semi-supervised approach to classify patients with Alzheimer’s disease from brain imaging, using the particle competition and cooperation model. In order to deal with images, feature extractors from general-purpose descriptors and pre-trained deep learning networks were used. The experiments were performed using the Open Access Series of Imaging Studies (OASIS) database. To evaluate the performance of the particle competition and cooperation model, its results were compared to those achieved by the SVM, kNN, Naive Bayes, Label Propagation and Label Spreading algorithms. The results have shown that this approach is promising and behaves as expected regarding the use of labeled data, i.e., the particle competition and cooperation model needs less labeled data, as it takes advantage of its semi-supervised learning approach.

Thanks to Prof. Dr. João Roberto Bertini Júnior (UNICAMP) and Prof. Dr. Danilo Medeiros Eler (UNESP) for their participation in the examination committee, and their valuable comments and suggestions.