Eduardo Pooch
Eduardo Pooch
About
I started my academic education with a BSc in Biomedical Informatics, obtained at the Federal University of Health Sciences of Porto Alegre (UFCSPA) in 2018. Then, I completed my MSc in Computer Science at the Pontifical Catholic University of Rio Grande do Sul (PUCRS) in 2021. During the MSc, I worked on the intersection of Machine Learning and Biomedical Imaging, with works on semi and self-supervised learning, domain generalisation, 3D object reconstruction, multimodal models and internships at Hewlett-Packard and Motorola. My thesis focused on pathology localisation on chest radiographs with limited labelled data through multiple-instance learning and I was awarded the Google Latin America Research Awards for it.
I am currently working towards a PhD at the Netherlands Cancer Institute (NKI-AVL) in Amsterdam, in collaboration with Maastricht University. My project focuses on identifying imaging biomarkers and developing deep learning models to aid the diagnosis of prostate cancer through imaging. Addressing clinical issues with AI simultaneously introduces technical challenges, including integrating multiple data modalities, handling data scarcity and ensuring model generalisation to unseen domains.