Artificial Intelligence for Better Patient Care

Hans-Ulrich Prokosch has known this for a long time. “Artificial intelligence can recognize patterns in vast amounts of data that a human simply could not,” says Prokosch, Professor for Medical Informatics at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) and CIO at Universitätsklinikum Erlangen. “AI algorithms provide us with valuable information for making decisions in preventative care, diagnostics, and treatment.”

Prokosch is one of the leading experts on the optimization of hospital information systems. These systems are used not only for administrative tasks but also for collecting patient data centrally—such as data on previous illnesses, treatment, medication, and much more. Hospital information systems are an important source of data that computer scientists need in order to develop innovative AI applications, which could be used to suggest suitable therapies and avoid unnecessary operations or repeated examinations.

Professor Prokosch and his team are exploring ways of using digital technology to connect other stakeholders in the healthcare system to ensure that these developments benefit all healthcare organisations rather than just specific hospitals. Establishing a data integration centre at Universitätsklinikum Erlangen is an important first step towards this goal: ‘We want to consolidate data from regional projects and healthcare services which would allow us to analyse the data in a structured way,’ explains Prokosch. This is already the case for the German Medical Informatics Initiative, which Prokosch is also involved in coordinating, for sharing data on COVID-19 patients between university hospitals in Germany.

But there is far more involved in establishing this charming university town in Frankonia as the “Bavarian AI Network Hub for Healthcare,” in which the Bavarian Government is investing millions of euros of research funding through the High-Tech Agenda Bavaria. In fact, the region around Erlangen and Nuremberg is home to an entire ecosystem of medical-science experts, companies, start-ups, and research institutions, making it an outstanding location for medical research.

Scientists have been researching artificial intelligence at FAU for over 40 years, long before the hype around AI began. FAU has a strong international reputation in medical engineering. Physicians at Universitätsklinikum Erlangen use the latest research findings in diagnostics and treatment.

Companies based in the region—from established global companies such as Siemens Healthineers to ambitious new start-ups—also benefit from knowledge and research transfer with FAU, ensuring that patients around the world benefit from expert research. Scientists from extramural research institutions such as the Max Planck Centre for Physics and Medicine are also exploring innovative approaches in the Bavarian AI Network Hub for Healthcare.

It is small wonder that many young researchers from around the world choose to start and develop their careers in the region—like Alessandro Del Vecchio, an expert in neuromuscular physiology and neural interfacing from Italy who has just been appointed to a professorship at FAU. Or Anne Koelewijn who specializes in biomechanical movement analysis and is one of the youngest professors at FAU.

Both professors conduct their research at the Department for Artificial Intelligence in Biomedical Engineering (AIBE) at FAU, a new department which will act as a nucleus for AI research at FAU, with 12 professors and around 100 research staff. The new department brings together many different fields in AI research.

Machine learning for diverse therapies, medical imaging and imaging analysis, remote health monitoring and telemedicine: research in interdisciplinary teams and clinical applications at FAU and Universitätsklinikum Erlangen deals with a broad range of topics. In various projects, researchers are testing AI-supported methods for health monitoring, for example: Scientists at the Chair of Digital Health have developed glasses that can record up to 100 different biomarkers—from heart rate to respiratory rate to food intake. Neurologists have developed a market-ready shoe that can monitor the gait of Parkinson’s patients using integrated sensors and provide feedback about their health. Researchers can use this data to provide objective behavioral and therapy recommendations.

Medical experts work closely with computer scientists at FAU to analyze and interpret data gathered from biomedical parameters or patient records. Prof. Dr. Björn Eskofier is the head of the AIBE department and Chair of Machine Learning and Data Analytics. He worked with FAU medical expert Prof. Dr. Jochen Klucken to develop the smart health-monitoring shoe. “An individual’s gait is as unique as their fingerprint. We therefore need self-learning algorithms that are capable of understanding patterns in an individual’s gait and monitoring changes,” explains Eskofier, who is an expert in pattern recognition using AI. “The more data we can collect and interpret meaningfully, the better the system will work.”

In the “SMART start” project, he is also collaborating with Prof. Dr. Matthias Beckmann and Prof. Dr. Peter Andreas Fasching from the Department of Obstetrics and Gynaecology at Universitätsklinikum Erlangen on improving healthcare services for pregnant patients. The project aims to develop a digital service to facilitate routine investigations.

A further example of AI-assisted diagnostics is interpreting data from medical imaging by analyzing x-ray, MRT, or CT images. FAU’s expertise in medical imaging is an important resource for physicians at Universitätsklinikum Erlangen. Prof. Dr. Andreas Maier develops algorithms for medical imaging applications at the Chair of Machine Intelligence. In the 4D+nanoSCOPE project, he is working on an innovative technology that will enable x-ray microscopy to be carried out on living subjects using computed tomography.

“Many people talk about interdisciplinary research but we live and breathe it here at FAU,” says Andreas Maier. “Medical experts identify approaches for improving healthcare, engineering and medical engineering experts develop new technology and devices, and computer scientists develop algorithms for analyzing imaging data and other data. We are in constant communication and provide each other with valuable input for our research on a daily basis.”

Complex challenges are best solved by many experts from many different fields. The Bavarian AI Network Hub for Healthcare enables deeply interdisciplinary collaboration through the world-leading medical engineering cluster Medical Valley, whose partners include the Faculty of Medicine, Faculty of Sciences, and Faculty of Engineering at FAU, Fraunhofer IIS, Universitätsklinikum Erlangen, and over 500 regional companies in the medical engineering and healthcare sector. This is an ideal environment for transferring innovation from research laboratories into market-ready products and practical applications.

Siemens Healthineers, a leading manufacturer of devices and solutions for medical imaging and laboratory diagnostics, is a founding partner of Medical Valley. “Findings and experience from clinical practice are an essential part of developing our technology and devices,” says Renate Jerecic, Head of Collaboration Office at Siemens Healthineers. ‘That is why we work directly with FAU and Universitätsklinikum Erlangen in many projects. Both organizations work together to share their innovative strengths.” Siemens Healthineers, FAU, and Universitätsklinikum Erlangen are currently working together on a project to develop one of the world’s most powerful MRT scanners with a field strength of seven tesla. The digital twin is also a collaboration between Siemens Healthineers, FAU, and Universitätsklinikum Erlangen.

It is not only global corporations such as Siemens Healthineers that are active in the cluster; ambitious young start-ups are also an important part of the network. Through Medical Valley they get the support they need to develop innovative health solutions: the start-up experts in Medical Valley, for example, arrange office space and contacts with strategic partners, as well as assisting with funding acquisition and public relations. Close collaboration and trusting partnerships are the lifeblood of the Digital Health Hub, whether between research institutions and companies or between individual departments. And they are essential to continuing its success story.