Fortunately, most diseases are well known, and their clinical picture is more or less clear. General practitioners or specialists can quickly make the right diagnosis for most of their patients or at least identify the disease by examining the patient and then starting treatment. But there are also rare diseases that even specialists find difficult to recognize.<\/p>\n\n\n\n
![\"\"](\"https:\/\/aktuelles.uni-frankfurt.de\/wp-content\/uploads\/2025\/03\/04_01_2.jpg\")
\u201cSome rare diseases are not so rare at all,\u201d says Michael von Wagner, Head of the Executive Department for Medical IT Systems and Digitalization and Chief Medical Informatics Officer (CMIO) at University Hospital Frankfurt. In Germany, around four million people are thought to have one of over 6,000 rare diseases. Some of these diseases are anything but exotic and by all means occur more frequently. \u201cIn the case of hepatitis C, for example, the liver parameters of up to 20 percent of patients is inconspicuous, but some of them show rheumatic symptoms,\u201d says von Wagner. \u201cAnd cystic fibrosis can progress in various ways and affect different organs.\u201d<\/p>\n\n\n\n
As the contact point for all kinds of complaints, general practitioners (GPs) are faced with the problem that it is impossible for them to have all these diseases on their radar. So where do they stand \u2013 especially if a patient comes back from the specialist and still does not have a clear diagnosis?<\/p>\n\n\n\n
\u201cFor many of these diseases, there are only a few specialists, even in a country like Germany, who are familiar with the course of the disease and the diagnosis and continue to conduct research into them,\u201d explains von Wagner. \u201cWe launched the SATURN project to ensure that general practitioners can find the right specialist if a diagnosis proves difficult.\u201d<\/p>\n\n\n\n
Smart portal for general practitioners<\/h3>\n\n\n\n
SATURN stands for Smartes Arztportal f\u00fcr Unklare Erkrankungen<\/em>(\u201cSmart Physician Portal for patients with unclear diseases\u201d). The project will run until the end of 2024 (www.saturn-projekt.de). In addition to University Hospital Frankfurt, other project partners include the Institute of General Medicine at Goethe University Frankfurt, the Institute for Medical Informatics and Biometry at TUD Dresden University of Technology, the Fraunhofer Institute for Experimental Software Engineering IESE in Kaiserslautern and the Institute of General Practice at the University of Cologne.<\/p>\n\n\n\n Various software-based diagnostic tools already exist, some of which integrate artificial intelligence \u2013 in other words, they use intelligent algorithms that are trained with large databases and can recognize important patterns in complex diseases. However, these diagnostic tools are difficult to use in practice.<\/p>\n\n\n\n This is mainly due to two problems. The first is the lack of specialists. Some systems suggest one diagnosis or another but still refer cases to specialists in different fields for further clarification. But such specialists rarely have time to respond to external inquiries during their daily work \u2013 and may even end up having to refer a case to a colleague. At the same time, general practitioners need a system that can help them find a suitable contact person as quickly as possible. In most cases, they only have time to search for specialists after they have finished their daily appointments. In this scenario, doctors really need a system that can direct them to the right specialist who, with a high degree of probability, will be able to help.<\/p>\n\n\n\n The second problem is that common machine learning methods are not suitable for medical questions. This is because such algorithms are generally designed to process large amounts of data and are good at recognizing recurring patterns in these data. \u201cIn medicine, however, it is often helpful to be able to find even one very similar case that has been correctly recognized and treated in the past,\u201d says von Wagner.<\/p>\n\n\n\n That is why the SATURN project is developing an artificial intelligence tool specifically tailored to medicine. It has three different components. The first is a rule-based decision-making system on the basis of medical guidelines and expert knowledge. Secondly, SATURN also uses a case database containing patients\u2019 well-documented medical data, symptoms and information about disease progression in anonymized form. And thirdly, specially customized machine learning algorithms that crawl the databases and prepare data for the attending physicians.<\/p>\n\n\n\n \u201cAt the end of the project, physicians will be able to upload a patient\u2019s symptoms and medical results onto a website,\u201d says von Wagner. \u201cThe system will then suggest a diagnosis and propose additional measures for common clinical pictures.\u201d The system will also be linked to the Health Care Atlas for People with Rare Diseases (SE-ATLAS), which lists experts and self-help organizations for patients with rare diseases.<\/p>\n\n\n\n \u201cWe want to make the entire system as practical and patient-friendly as possible,\u201d explains von Wagner, \u201cwhich is why we have worked closely with physicians from the outset.\u201d To ensure that they actually use the system, it must be adapted to their requirements and workflows.<\/p>\n\n\n\n If a GP is confronted with a difficult case that they cannot identify, the smart portal will in the future help them in several different ways. First, the system compares the patient\u2019s symptoms with findings in the database from clinically validated cases. It then proposes a diagnosis based on these data. \u201cThis does not have to be correct and should be clarified with a specialist if necessary,\u201d says von Wagner. Drawing on data from SE-ATLAS, the system also gives the attending physician the right contact information.<\/p>\n\n\n\n Even for experienced physicians it is not always easy and often time-consuming to find the right contact person for rare diseases. And even an expert cannot keep track of every rare disease that manifests itself in one way or another in their specialist field.<\/p>\n\n\n\n That is why SATURN links the results in the database with appropriate contacts to experts and self-help organizations. \u201cIt is precisely this combination of diagnostic support with relevant contact information that the medical profession and patient representatives want and which we hope will lead to real progress in the treatment of rare diseases,\u201d says von Wagner. So far, no expert system has brought these requirements together like SATURN does. By the end of the project, a pilot version will have been developed that can be evaluated and find application as a smart portal for physicians. The system would then have to be approved.<\/p>\n\n\n\n Also included in the specifications are cybersecurity and data protection. \u201cWe don\u2019t see any obstacles here because we have designed the system in such a way that all these requirements can be met or are already met,\u201d explains von Wagner. \u201cWe would like to see our smart portal integrated into existing software already used by physicians.\u201d This would have the advantage that GPs could search directly in the software without having to open their browser and type in the patient\u2019s symptoms. This would simplify workflow and save a certain amount of time.<\/p>\n\n\n\n But developing the portal is currently the top priority. A first test version went online at the end of February 2024. The project team hopes for extensive and critical feedback to ensure that the portal meets the needs of its users. \u201cI am very optimistic that we can make real progress in the diagnosis of rare diseases with this project,\u201d concludes von Wagner. \u201cWe have excellent clinical data, a suitable AI algorithm and software adapted to clinical practice.\u201d And in rural areas, where access to specialists is often difficult, such a system could be of great help.<\/p>\n\n\n\n The DocCheck<\/a> website lists 177 of the 17,000 known rare diseases<\/p>\n\n\n\n About \/ Michael von Wagner<\/strong>, born in 1972, holds a doctoral degree in medicine and is Head of the Executive Department for Medical IT Systems and Digitalization, Office of the Medical Director and Chair of the Executive Board, and Chief Medical Informatics Officer (CMIO) at University Hospital Frankfurt. He works closely with Jens Schulze, Chief Information Officer, and Professor Holger Storf, Director of the Institute of Medical Informatics, and is Managing Director of the joint University Center for Digital Healthcare. He is responsible for clinical IT systems and the strategic development of digitalization. The author \/ Dirk Eidem\u00fcller<\/strong>, born in 1975, studied physics (major) and philosophy (minor) in Darmstadt, Heidelberg, Rome and Berlin, graduated in astroparticle physics and holds a doctoral degree in philosophy of science. He lives in Berlin and works as a freelance author and science journalist.Fit for clinical practice<\/h3>\n\n\n\n
![\"\"](\"https:\/\/aktuelles.uni-frankfurt.de\/wp-content\/uploads\/2025\/03\/Person_Wagnervon-500x500.jpg\")
michael.wagner@ukffm.de<\/a><\/p>\n\n\n\n![\"\"](\"https:\/\/aktuelles.uni-frankfurt.de\/wp-content\/uploads\/2025\/03\/Autor_Eidemueller-500x500.jpg\")
dirk.eidemueller@gmx.de<\/a><\/p>\n\n\n\n