Hi, I’m Anna.
Computer vision & generative AI.
I’m a PhD researcher working on diffusion models for medical imaging, exploring how AI can understand and generate visual information.
Bio.
My background spans signal processing, machine learning and deep learning, with experience in 3D reconstruction and document understanding. I hold a Master’s in Computer Vision (UAB) and a Bachelor’s in Telecommunications Engineering (UPC), and have worked in industry and research in Barcelona and Lausanne, including LTS5 at EPFL.
Today, I’m a researcher in Eurecat’s Digital Health Unit and a PhD candidate with the AIBA group at Universitat de Barcelona, supervised by Prof. Petia Radeva and Dr. Vicent Ribas. My current work explores generative models and visual synthesis, applied to medical imaging.
Explore a model
in action.
MUMINS is a diffusion model that predicts a future medical scan and estimates its uncertainty from a baseline scan.
Explore a MUMINS sample: switch between brain MRI and lung CT, move through the slices, and compare each prediction with its uncertainty map.
About the MUMINS model ↗
Selected publications.
MUMINS: Metadata-conditioned Uncertainty-aware Medical Image Next-state Synthesis
IEEE/CVF Winter Conference on Applications of Computer Vision
A diffusion framework that synthesizes a future 3D scan from a baseline scan and a target time interval, producing a voxel-wise uncertainty map in the same sampling process. Evaluated on lung CT and brain MRI.
Anatomically Guided Latent Diffusion for High-Resolution 3D Chest CT Synthesis
Scientific Reports by Nature
A latent diffusion model that generates high-resolution 3D chest CT volumes conditioned on anatomical masks — producing realistic, high-quality synthetic scans at a fraction of the usual computational cost.
Out of the lab.
Phase IV-AI LAND: 3D Thoracic CT Synthesis with Anatomical Guidance via Lung and Nodule Diffusion
CIDAI - Centre of Innovation for Data Tech and Artificial Intelligence · Presentation of Proofs of Concept in Data and AI
Invited talk presenting LAND's proof-of-concept.
Pre-processing for Automatic Cervical Cancer Screening: Image Segmentation and Movement Detection
MICAD 3rd Conference on Medical Imaging and Computer-Aided Diagnosis
Abstract oral presentation of the bachelor's thesis work carried out at LTS5, EPFL.
Phase IV-AI LAND: 3D Thoracic CT Synthesis with Anatomical Guidance via Lung and Nodule Diffusion
Phase IV-AI Project Ending Conference
Poster presenting LAND at the Phase IV-AI project's closing conference.
LAND: 3D Thoracic CT Synthesis with Anatomical Guidance via Lung and Nodule Diffusion
MedEurIPS Workshop: Medical Imaging meets EurIPS
Poster session presenting LAND alongside the oral paper presentation.
Let’s connect.
↗Interested in computer vision, generative models, or a research collaboration? Get in touch.