基于物理信息的动态心脏磁共振成像生成式AI博士研究员

Do you want to make cardiac MRI safer for patients and the environment? Join PRIMA-CMR, a EuroTech Universities Alliance project between TU/e and the Technical University of Munich (TUM) and help to develop trustworthy generative AI that makes it possible to scan with far less contrast agent.

Information Cardiac MRI gives unmatched insight into heart disease, but some of its most important exams depend on gadolinium-based contrast agents. Gadolinium can build up in the body, cannot be given to all patients, and is now an environmental pollutant. A large reduction in the dose would make cardiac MRI safer and available to more patients. At low doses, however, the images no longer contain enough signal for a reliable diagnosis. Current AI methods can make images that look realistic, but they cannot guarantee that what they show is real.

We are looking for a motivated PhD candidate to develop new generative AI methods that recover diagnostic information from low-dose cardiac MRI in a reliable and trustworthy way. Your work will combine knowledge of MR imaging physics, patient-specific clinical information, and uncertainty quantification. The aim is a model whose outputs clinicians can trust, and which tells them when it is not certain. You will validate the methods on real clinical data together with our clinical partners.

This PhD position is part of PRIMA-CMR, a collaborative project within the EuroTech Universities Alliance programme. You will be based at TU/e and spend six months at TUM in the group of prof. Julia Schnabel, which is a world leader in generative and multimodal AI. You will also work with international clinical and academic partners across Europe. This position is embedded in the Medical Image Analysis group, Department of Biomedical Engineering, TU/e. You will be supervised by dr. Cian Scannell, and co-supervised by prof. Julia Schnabel (TUM).