Beatriz Neto is a Master’s student in Electrical and Computer Engineering at Instituto Superior Técnico, majoring in Control, Robotics, and Artificial Intelligence. In affiliation with the Champalimaud Foundation, she is developing her thesis on advancing MRI acquisition and reconstruction through state-of-the-art generative AI methods.
Her research focuses on denoising diffusion models as a flexible framework for enhancing low-SNR MRI data, particularly diffusion-weighted imaging (DWI). By adapting these models to recover the quality of multi-acquisition scans from single-acquisition data, her work aims to reduce scan times, mitigate motion artifacts, and improve patient comfort without compromising diagnostic value. Through quantitative benchmarking and clinical evaluation, her study seeks to demonstrate the potential of deep generative models to make MRI more efficient and accessible.