BINDER - Improving Bio-inspired Deep Learning for Radiomics
Description
The project's objective is to improve the state of the art in Radiomics analysis of breast and rectal cancer, using existing and novel Machine Learning (ML) and Deep Learning (DL) methods. Radiomics is an emerging field of oncologic imaging. It aims at extracting large amounts of informative features from standard of care medical images, and analysing them for improving predictive power in precision medicine. Currently, many phases of the Radiomics workflow are not automated, and thus time-consuming, subjective, and prone to errors. The project will contribute at the development of faster, more objective, and accurate Radiomics models, that can be used to tailor treatment options and thus reduce toxicity and improve clinical outcomes. To achieve this goal, among others, Convolutional NNs will be used and compared to novel DL methods, that are expected to outperform the state of the art, thanks to their competitiveness in terms of evolvability and their ability of limiting overfitting.
Researchers
Funded by
P2020
Partners
Universidade Nova de Lisboa, Universidade de Lisboa, Universidade de Coimbra, Fundação Champalimaud
Total budget
239 696,00 €
Local budget
13 518,00 €
Keywords
Radiomics, Machine Learning, Deep Learning, Genetic Programming
Start Date
2019-01-01
End Date
2021-12-31