This project will be focused on the development of supervised and unsupervised machine learning algorithms for the processing of cardiac magnetic resonance images, with the objective of differentiating different types of hypertrophic cardiomyopathies. Initially, segmentation algorithms based on deep learning architectures will be developed, based on state-of-the-art neural networks (e.g., U-net variations, transformers). Consequently, techniques such as multiple kernel learning will be used to create dimensionality reduced latent spaces to identify different patient phenotypes. The project will be performed at Hospital de la Santa Creu I Sant Pau, thus the developed computational methodologies need to be adapted to work in a clinical environment. 

Supervisors: Martín Descalzo, Matías Calandrelli (Hospital de Sant Pau),  Oscar Camara