Marina Crespo Aguirre

Biomedical Engineer

About Me

Hey there ! I am a Biomedical Engineer with great passion for the intersection of machine learning and medical imaging. I am specially interested in the topics of generative modeling and self-supervised, low-budget machine learning.

I am currently pursuing my PhD at the University of Basel, where I am part of the Analytics and Informatics for Child Health Group (AICH) . Previously, I completed my MSc Degree in Biomedical Imaging at ETH Zürich, supported by La Caixa Foundation.

Throughout my studies, I have collaborated with research groups at the Universitat Politècnica de València (BioITACA), the Paul Scherrer Institute, ETH Zürich (BMIC group), and Harvard Medical School (LEMoN group).

Feel free to explore some of my past projects! 🚀

Publications

Improving Neuropathological Reconstruction Fidelity via AI Slice Imputation

Marina Crespo Aguirre, Jonathan Williams-Ramirez, Dina Zemlyanker, Xiaoling Hu, Lucas J. Deden-Binder, Rogeny Herisse, Mark Montine, Theresa R. Connors, Christopher Mount, Christine L. MacDonald, C. Dirk Keene, Caitlin S. Latimer, Derek H. Oakley, Bradley T. Hyman, Ana Lawry Aguila Juan Eugenio Iglesias

HTML GIT

Evaluation and assessment of clique arrangements for the estimation of omnipolar electrograms in high density electrode arrays: An experimental animal model study

Samuel Ruipérez-Campillo, Marina Crespo, Álvaro Tormos, Antonio Guill, Antonio Cebrián, Antonio Alberola, Jakob Heimer, Francisco J Chorro, José Millet and Francisco Castells

Physical and Engineering Sciences in Medicine, 2023

HTML GIT

Conferences

Assessment of the Interelectrode Distance Effect over the Omnipole with High Multielectrode Arrays

Marina Crespo, S Ruipérez Campillo, Rubén Casado-Arroyo, José Millet and Francisco Castells

45th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2023

HTML

Estudio Comparativo con Señales Epicárdicas de las Limitaciones del Omnipolo con Multielectrodos de Alta Densidad

M Crespo Aguirre, S Ruipérez Campillo, I Segarra, A Guill, A Tormos, J Millet and F Castells

XL Congreso Anual de la Sociedad Española de Ingeniería Biomédica, 2022

HTML

Featured Projects

Neural Implicit Fields for the reconstruction of fastMRI

Neural Implicit Fields for reconstructing MRI scans from undersampled, aliased data. The dataset used for this project belongs to the open-source fastMRI dataset. The repository contains different training pipelines, regularization losses and input encoding strategies. The research forms part of a Semester Project carried out at CVL (Fall 2024).

GIT

Medical Image Generation, Retraining on a Strict Budget

Low-computation approach for Data Augmentation using the formulation proposed in SeLoRA. We implemented the methodology to two datasets ( BraTS and PI-CAI), and validated the outputs on a downstream classifier. We achieved comparatively better AUC scores than other baseline methods (i.e. GANS, LoRA). This work belongs to an assignment from the course Deep Learning, ETH Zürich (Fall 2024).

GIT