30/10/2026
🔬Can 3D breast imaging and artificial intelligence help to better characterize breast cancer in a noninvasive way?
Identifying breast cancer molecular subtypes is crucial for guiding treatment decisions.
A new study published in Biomedicines explores the use of 3D convolutional neural networks (CNNs) to predict molecular subtypes directly from 3D breast tomosynthesis images, investigating how different volumetric representations influence prediction performance.
The research was conducted by Raquel Tavares, as part of her Master's research in Biomedical Engineering and Biophysics at Faculdade de Ciências da Universidade de Lisboa, under the supervision of Ana Margarida Mota (Instituto de Biofísica e Engenharia Biomédica - Institute of Biophysics and Biomedical Engineering), in collaboration with Eunice Carrasquinha (Centro de Estatística e Aplicações da Universidade de Lisboa).
🔗 Read more at https://www.mdpi.com/2227-9059/14/11/2411