Healthcare systems increasingly generate data that can be used for such purposes as advancing knowledge on diagnostics and combatting disease. These health data are of different types, including patient records of scanning images and genetic data. These data have great potential if we can understand how to collect and analyse them correctly.
Achieving this requires having appropriate research infrastructure. The Novo Nordisk Foundation has therefore awarded DKK 27.9 million for two projects that will develop advanced infrastructure to strengthen research on health both nationally and internationally.
Protected training environment for health data and improving mapping of the brain
One project, the National Health Data Science Sandbox for Training and Research, will be based at the Center for Health Data Science, Faculty of Health and Medical Sciences, University of Copenhagen. Researchers throughout Denmark will collaborate on building a national Sandbox infrastructure that will provide students and researchers with access to synthetic health data in a specially designed environment. The new infrastructure is a collaborative effort among researchers across Denmark and will facilitate the improved education and training of a new generation of clinical data scientists in Denmark.
Users of the Sandbox will have access to the actual data and anonymized synthetic data and tools, computer programmes and computing power from supercomputers, similar to working on a project with the actual health data. The users will instead find themselves in the protected test area of the Sandbox that can ease the transition to working with the actual health data.
“Today, a closed environment is lacking in which new ideas can be tested without needing to use time for the application procedures required to work with personally sensitive health data. The new Sandbox infrastructure will give students and researchers the opportunity to train on processing health data, whereas access to the actual health data can be reserved for the most promising projects, for which investing more time and resources make sense,” says grant recipient Anders Krogh, Professor, Center for Health Data Science, University of Copenhagen.
The purpose of the second project, The OpenNeuroPET Archive – A Molecular Neuroimaging Archive, is to establish an open-access platform that can systematically and uniformly process and archive data from brain imaging. This database will enable researchers worldwide to share data to advance brain research and medical imaging technology. It will be specifically designed so that the data sharing complies with the General Data Protection Regulation of the European Union.
The new infrastructure will enable researchers to collect and reuse data from neuroimaging carried out using positron emission tomography (PET), a method for measuring the quantity and location of specific molecules in the brain.
Since PET neuroimaging is expensive and requires exposure to ionizing radiation, studies typically include a limited number of patients and healthy control participants. By collecting and sharing data from the PET neuroimaging internationally, across institutions, the new platform will enable large and expensive quantities of acquired data to be collected in the archive, and more researchers can have access to analyse these huge quantities of data. One application for this is to accurately map certain molecules in the brain, thereby producing an atlas of the brain.
“More solid research results require the sharing of data. We need to get used to doing this and thereby being able to confirm the results from small studies and to refine the conclusions. This will help us to find more certain answers to important research questions,” explains grant recipient Gitte Moos Knudsen, Professor, Neurobiology Research Unit, Rigshospitalet, Copenhagen, who is building the new platform in collaboration with researchers from the Department of Computer Science at the University of Copenhagen and others.
Both projects will ensure that the infrastructure is freely accessible to other researchers to maximize the benefits. The project grants have been awarded in open competition through the Foundation’s Data Science Research Infrastructure Programme and are two of the 12 projects receiving grants in 2020 through the Foundation’s new Data Science Initiative. The Foundation has allocated DKK 410 million to the initiative in 2020–2022. The applications for the 2021 programmes open at the end of December 2020.
“Access to large quantities of data are the basis of the breakthroughs we are currently experiencing within artificial intelligence, machine learning and the development of advanced models. The Foundation’s grants for research on data science therefore naturally parallel grants for infrastructure that can support this research by creating access to large, unique data sets that are relevant to the healthcare sector but also in research in such areas as biotechnology, sustainability and research in the natural and technical sciences,” says Niels-Henrik von Holstein-Rathlou, Senior Vice President, Biomedicine & Health Sciences, Novo Nordisk Foundation.
Data Science Research Infrastructure Programme grants in 2020
National Health Data Science Sandbox for Training and Research, Anders Krogh, Professor, Center for Health Data Science, Faculty of Health and Medical Sciences, University of Copenhagen: DKK 17,764,483
The OpenNeuroPET Archive – A Molecular Neuroimaging Archive, Gitte Moos Knudsen, Professor, Neurobiology Research Unit, Rigshospitalet, Copenhagen: DKK 10,144,473
Sabina Askholm Larsen, Communications Partner, +45 2367 3226, [email protected]
Anders Krogh, Professor, University of Copenhagen, +45 5182 7056, [email protected]
Gitte Moos Knudsen, Professor, Rigshospitalet, +45 4063 0305, [email protected]