My research aims at learning feature representations for medical image diagnosis facing common medical imaging data challenges. Namely limited data, class-imbalance, noisy annotations and data privacy for collaborative learning between different hospitals.
I have explored the role of architectural design in dealing with spatial information for several classification tasks. Also, I have investigated easing the optimization of the deep network parameters by gradually increasing the difficulty of the training samples. This gradual increase is based on the concept of curriculum learning and achieved with a data scheduler that controls the order and pace of the samples.
Below there is a summary of the of research projects I have been working on:
Students Cathrine Damgaard & Trine Naja Eriksen (ITU) - Detecting Shortcuts in Chest X-ray Images (Research Project) 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)