Research

I am a data scientist working as a postdoctoral researcher at the University of Barcelona. My research focuses on data-centric approaches for fostering responsible machine learning in healthcare applications. Previously, I was a postdoctoral researcher at the IT University of Copenhagen (ITU), under the supervision of Prof. Veronika Cheplygina. I earned my PhD with “Cum Laude” from Pompeu Fabra University (UPF), under the supervision of Prof. Gemma Piella (UPF) and Prof. Diana Mateus (École Centrale Nantes).

During my PhD at UPF, I investigated training sample strategies, addressing common challenges in medical imaging such limited data, class-imbalance, and noisy labels. I developed curriculum learning strategies that leveraged prior clinical knowledge from medical guidelines or estimated model uncertainty. In my postdoctoral research at ITU, I investigated transfer learning, with an emphasis on identifying proper pre-training sources and evaluating transferability metrics to improve model robustness. I also gained experience in conducting fairness analysis and critical data studies approaching these topics from a socio-technical perspective.

Keywords: medical imaging, transfer learning, federated learning, domain adaptation, classification, limited data, noisy labels, critical data studies, data governance.

Projects

During my career, I have pursued various lines of research, which I summarize below:

Students

  • Nikolaj M. Sømod (ITU) - Data Representativity, Similarity, and Diversity (Master Thesis)
  • Bianca Ida Pedersen & Max Andreas de Visser (ITU) - Feature Dependency in Deep Ordinal Regression Networks for Lung Nodule Malignancy Prediction (Master Thesis)
  • Patrick Wittendorff Abarzua Neira, Silas Roien Arildsen, Bertram Kosmo Hviid & Tobias Skovbæk Brund (ITU) - Machine Learning for Sign Language Recognition (Bachelor Thesis)
  • Paula Victoria Menshikoff & Katarina Kraljevic (ITU) - Diffusion-based Shortcut Removal in Chest X-ray Datasets (Bachelor Thesis)
  • Casper Anton Poulsen (ITU) - Multi-Task Learning for Pathology and Tube Detection (Research Project)
  • Cathrine Damgaard & Trine Naja Eriksen (ITU) - Annotation Label Reliability, Drain and Pathology Detection in Publicly Available Chest X-ray Datasets (Research Project + Master Thesis)
  • Olalla Aramburu (UPF) - Enhancing Surgeon Action Detection in Robot-Assisted Minimally Invasive Surgery (Bachelor Thesis)
  • Joan Medina (UPF) - Predicting Intracranial Pressure with Recurrent and Domain Adaptation Neural Networks (Bachelor Thesis)
  • Aswathi (EC Nantes) - COVID-19 Detection with a Scheduled Convolutional Neural Network (Research Project)
  • Eloi Francisco (UPF) - Training Deep Neural Networks on Noisy Labels with Bootstrapping (Research Project)
  • Simran Anand (UPF) - Identifying Late Gadolinium Enhancement in Cardiac Magnetic Resonance Images (Bachelor Thesis)
  • Domingo de Abreu (UPF) - Musculoskeletal Abnormality Detection on X-ray Using Transfer Learning (Master Thesis)
  • Kami Artik (UPF) - Distilling Knowledge in Convolutional Neural Networks to Detect the Abnormalities in Radiographs (Master Thesis)