Large AI project receives over €95 million for ten years of public-private research

The Dutch Research Council (NWO) is to fund two consortia for a period of ten years. Radboudumc participates in the ROBUST consortium, which consists of 17 AI labs, eight of which are dedicated to healthcare. Radboudumc leads five of these eight labs. "A unique opportunity to strengthen our leading international position in AI for healthcare in Nijmegen," says Bram van Ginneken, professor of AI at Radboudumc.

Over the next ten years, the ROBUST consortium, coordinated by Maarten de Rijke (UvA/ICAI), will carry out research into reliable artificial intelligence (AI) with a total budget of more than €95 million. It is a large public-private programme involving 54 partners (21 knowledge institutions, including four universities of applied sciences, 23 companies and ten civil society organisations) that researches the development and application of AI in a wide range of fields, such as the energy, railway, media, high-tech and healthcare sectors.

Healthcare is particularly well represented in the consortium, accounting for eight out of a total of 17 AI labs. Radboudumc leads five of those eight labs. A sixth lab will be set up at Radboud University, meaning that a third of the project will be based in Nijmegen. Each lab will receive ten PhD candidates, a permanent member of staff and support from scientific programmers. The research conducted at each lab will also address ethical and social issues.

Continuation

"For Radboud university medical center specifically, this means that two existing ICAI labs will be expanded," says AI professor Bram van Ginneken. "One of those labs is the Thira lab, which we founded a few years ago in collaboration with Thirona – a spin-off between Radboud university medical center and Delft Imaging. The ten PhD candidates will help us to develop new AI applications not only for lung diseases but also to facilitate pregnancy ultrasounds in developing countries, for example."

"AI for Health is the other lab in which we have been collaborating with Radboud University for a while now. One of the new partners that will join this lab is Verily, which is a part of Google. Together we will investigate how we can use AI to generate more knowledge about and improve care for people with Parkinson’s disease."

Three new labs

Radboud university medical center’s Cardiology Department is joining forces with Abbott to investigate AI for optical coherence tomography (OCT). Van Ginneken: "It is a fairly new technique that improves the imaging of the heart’s coronary arteries. But it generates mountains of data. We want to use AI to predict which patients are at high risk of a new infarction so that this group can be given extra medication. We are going to be working with Siemens Healthineers to investigate how MRI images can be used even more intelligently, and how we can use AI to extract even more information from them."

"We are setting up the third new lab in collaboration with MeVis Medical Solutions. They are based in Germany, and we’ve been working with them for a very long time. Together we are going to further develop the software for CT lung cancer screening, which was developed at Radboud university medical center and is now used all over the world, and potentially also apply it to other medical fields."

"It is fantastic that we now have guaranteed funding for the next ten years. This gives us the opportunity to develop the next generation of AI systems for healthcare with a large group of researchers based at Radboud university medical center, and to actually start using them in practice in collaboration with our partners from industry."

Medical AI

The ROBUST consortium (Trustworthy AI-based systems for sustainable growth) is a public-private project, which is largely funded by the participating companies. "For many fields, it is not easy to obtain a financial contribution of this size from industry," says Van Ginneken. "But any company that wants to introduce medical AI technology to the market has to work with hospitals because of the need for anonymised patient data and the crucial input from doctors that is required for this. The algorithms we create are constantly being improved, retrained and validated."

Nijmegen’s AI ecosystem

The granting of funding for the ROBUST project represents a major boost for AI research at Radboud university medical center. The medical centre already has a lot of connections with companies, which are eager to cooperate with Radboud university medical center because of its good track record. There is also scope to generate a real impact on healthcare through spin-offs such as Thirona and Screenpoint Medical. "The interconnectedness of world-class research at Radboudumc and Radboud University, combined with the strength of the companies, gives us the opportunity to build a strong and thriving AI ecosystem," says van Ginneken.

Innovation and Top Sectors

A long-term programme (KIC) is a powerful catalyst for the development of a scientific field, geared towards a specific societal theme or key technology. Through this programme, NWO offers long-term funding for structural cooperation between public and private parties and knowledge institutions. NWO earmarks approximately €80 million for this purpose over a period of four years.

These long-term programmes constitute one of NWO’s main strategic contributions to the Knowledge and Innovation Covenant 2020-2023 (KIC). The KIC research programme is dedicated to ground-breaking innovative solutions with social and economic impact. NWO connects companies and knowledge institutions and funds ground-breaking research based on their innovative, high-impact research proposals. This is laid down in the Knowledge and Innovation Covenant (KIC) 2020-2023, which ties in with the central government’s mission-driven top sectors and innovation policy.

• Want to know more about these subjects? Click on the buttons below for more news.

  Digital Research EnvironmentNews homepage news & media