DLMedIA: Deep generative models for cardiovascular disease

Developing deep generative models to learn more efficiently from less data in the analysis of cardiac CT.

ONGOING

2017

 

 | 

2022

 

RESEARCH AREAS

About project

Deep learning has been very successful for image analysis when a very large amount of data is available. Due to privacy legislation and expert knowledge required for annotation, it is much harder to accumulate a similar number of medical images for training deep networks. The experience has been that on much smaller image datasets deep convolutional neural networks are difficult to train reliably. In this project, we aim to incorporate expert knowledge in the form of generative models in deep learning in order to learn more efficiently from less data. In our group, we focus on the development of deep learning models for analysis of CT images of the heart. The project is part of the TTW Perspectief DLMedia project and partially funded by Philips Healthcare.

 

Our publications & software