DZNE Magdeburg develops digital "red-yellow-green disclosure" for MRI scans

Magnetic resonance imaging (MRI) is an imaging technique used to show organs and tissue inside the body. However, if MRI images are blurred, these images are inadequate - and thus useless - for clinical care and for conducting medical studies. Prof. Emrah Düzel, site speaker and head of Clinical Research at the DZNE Magdeburg, together with his research group and experts from the University Medical Center Magdeburg, has now developed an artificial intelligence to digitally evaluate and improve the image quality of MRI images at the IT event "HealthCare Hackathon".

Blurry images or blurred structures: During magnetic resonance imaging (MRI), so-called image artifacts can quickly occur - for example, if the person being examined moves during the image acquisition. Normally, the image quality must be assessed by specialists in clinical routine. If image errors occur or the person making the assessment is inexperienced and uncertain in the evaluation, it is necessary to repeat the examination. This is also the case if the time is too short for an evaluation due to a lack of personnel and it is only discovered after the examination date that the image quality is insufficient. At the software and hardware development event "HealthCare Hackathon" in Mainz, Emrah Düzel's research group, together with employees of the University Department of Neurology and the IT and the Medical Computer Center of University Medical Center Magdeburg, has now developed an artificial intelligence that can recognize artifacts in MRI image data and evaluate their severity. At the "HealthCare Hackathon" interdisciplinary teams jointly develop IT projects for problems in the fields of nursing, artificial intelligence, quantum computing, bots or emergency medicine.

The Artificial Intelligence created by the Magdeburg team is to provide a real-time red-yellow-green disclosure for the image quality (usable/questionable/unusuable) of MRI images: In case of unusuable scans, the image must be retaken immediately. Scans with questionable image quality will be reviewed again by qualified personnel. Usable scans are sent directly to a doctor for diagnosis. The aim of this digital solution is to identify recurring errors and to set up a completely automated evaluation. This saves costs and, above all, time, because scans do not have to be repeated and processes and procedures can be optimized. The team won an implementation workshop for this project with Siemens Healthcare, IBM Germany and Amazon Web Services DACH and was invited to the "Healthcare Hackathon" 2021 in Berlin.

In the future, not only the clinics of the University Medical Center Magdeburg should benefit from the project, but also smaller hospitals and radiological practices in Saxony-Anhalt. The automated assessment red-yellow-green disclosure can also be used in clinical research studies of the DZNE to support quality assurance for studies.


The "Healthcare Hackathon" is an interactive and interdisciplinary event format that has been taking place in various cities since 2016. IT professionals meet other interested participants to develop new tools and processes in teams, which are applied by many industrial partners and in university medicine.

The team of employees of the DZNE Magdeburg, the University Department of Neurology and the Medical Computer Center of University Medical Center Magdeburg participated for the first time in a "Healthcare Hackathon". Experts from 23 university hospitals took part in the event.

Welcome to our website, here you can inform yourself basically cookie-free.

We would be pleased if you would allow a cookie to be set for analysis purposes in order to optimise our provided information. All data are pseudonymous and are only used by the DZNE. We deliberately avoid third-party cookies. You can deselect this setting at any time here.

Your browser allows the setting of cookies: