Convolutional Deep Learning for multimodal classification of abnormal electrical brain activity

Using medical image data, this thesis will focus on the adoption of deep learning models for harmful brain activity classification. In particular, a detection and classification of anomalies related to electroencephalograms (EEGs) will be performed. Convolutional neural networks (CNNs) will be trained and fine-tuned on a dataset provided by a medical centre, with the aim of improving the identification of pathological brain patterns and contributing to faster and more accurate diagnoses.