RWTH: Digitized data acquisition and AI for safe operations

KIPeriOP is a research project funded by the Federal Ministry of Health with 1.5 million euros until September 2023. The project is coordinated by Professor Anja Hennemuth from the Fraunhofer Institute for Digital Medicine MEVIS and Professor Patrick Meybohm from the University Hospital Würzburg. Doctors from Asklepios Medical School GmbH, the Frankfurt University Hospital and the Charité University Medicine Berlin are involved. You will work with specialists from the fields of AI, user guidance, ethics and health economics, including Professor Saskia K. Nagel, head of the teaching and research area Applied Ethics in the department for Society, Technology and Human Factors of Faculty 7 at RWTH Aachen University.

Every year, more than 16 million operations are performed in Germany. There are always complications that often lead to death: In the western industrialized nations, 0.4 to 0.8 percent of those who have been operated on die during or after an operation. The clinics try to reduce this number, among other things, by taking possible risk factors into account: What comorbidities does a patient have, what medication is currently being taken? What complications could this cause and how can they be minimized?

Although there are guideline papers that support medical staff in this risk assessment, they list the type and number of useful preliminary examinations. However, in practice these guidelines are not easily applicable. They are complex documents and their application requires consideration of a wealth of information that is not always easy to obtain.

This is where the “KIPeriOP” research project comes in, carried out by an interdisciplinary consortium. The aim is to develop a clinical decision support system, known in technical jargon as the CDS system. The software developed by Börm-Bruckmeier Verlag should first collect possible risk factors for each individual patient and in accordance with guidelines, put them in relation to one another and, as a result, provide a risk assessment: How likely is it for a particular patient that serious complications will occur during or after an operation? “On the basis of this risk assessment, doctors can, for example, decide whether further examinations are necessary and which measures can be used to optimally prepare the patient for the operation,” explains Meybohm.

As input data as possible, as much information as possible about the respective patient should flow into the CDS system, including laboratory values, medication plan, vital data and information about lifestyle habits. In addition to taking the guideline into account, Artificial Intelligence (AI) will also analyze the digitally recorded data in KIPeriOP: Learning algorithms look for patterns and correlations that reveal which constellations of risk factors are likely to lead to which complications. With the help of the AI, it could be better recognized that a patient suffers from an undetected cardiac insufficiency and thus has an increased risk of surgery.

Various AI processes are being tested in the project in order to find an optimal model. In order for them to function reliably, the algorithms must first be trained, ie fed with a large number of data records about actual preliminary examinations and the course of the operation. The four clinical project partners collect this data. “We not only collect data here that are already available anyway, but can also adapt the data collection to our needs,” says Meybohm.

One of the challenges in developing the CDS system is its usability. “We have to design the AI-based solution in such a way that it enriches the work of doctors and is not perceived as a burden,” emphasizes Hennemuth. “The necessary trust in the new technology can only arise if we make it transparent how and with what certainty the algorithms arrive at their results.”

Accordingly, the AI ​​should not act as a black box, but also make uncertainties and possible sources of error transparent. The development takes place in close coordination with the clinical partners, but also with the ethics team at RWTH Aachen University.

Comments are closed.