In everyday life, algorithms already support us in many activities, and in healthcare, too, artificial intelligence (AI) is considered a key technology and driver for progress and development. Classic treatment paths but also diagnostics will change and the use of AI will play a crucial role in this.
© iStock/Amnaj Khetsamtip
Together with Flying Health, a company that brings together leading players in the healthcare industry, the Fraunhofer Institute for Cognitive Systems IKS hosted the AI Innovation Days in Berlin. Here, people from different areas of healthcare collaborated with experts in reliable AI to jointly develop AI-based solution approaches for practical use cases for the benefit of patients. Experts from all areas of healthcare participated, from hospitals, health insurers, and pharmaceutical companies to medical technology and technology service providers.
For three days, working groups pored over their ideas, discussed applications of AI in healthcare, and developed approaches to solutions, resulting in four highly promising project outlines.
The AI-based prediction tool CoRe aims to improve hospital processes and apply guideline-compliant measures preventively to be able to react to complications in the clinic in advance and minimize the risk of readmission. The Risk Assessment Tool is designed to predict, prior to surgery, whether a person to be operated on is at high risk of certain complications based on the evaluation of various data. With this support, certain measures can already be considered during the treatment.
In addition, the experts developed a solution for predictive workforce management to make more accurate and automated predictions about staffing requirements in hospital emergency departments. The aim is to avoid critical overload situations by making predictions and planning staffing accordingly. This, in turn, leads to higher staff satisfaction, which in turn improves treatment quality and reduces costs.
To simplify physician-patient interactions and thus achieve data-driven healthcare, another group worked on the Easy Docu project idea, a tool for automated documentation using Natural Language Processing (NLP). This leaves more time for the patient in doctor-patient conversations and makes communication more personal. Connected assessments and continuous transparency of documentation throughout the entire patient pathway also enable a better understanding of treatment and thus high-quality therapy. Patient safety is also improved due to a reduced susceptibility to errors.
Another project related to AI-based support for personalized therapy selection. Using historical data on a patient, the aim is to make an AI-based prediction of how likely a complex therapy is to be adhered to (adherence). This allows the physician to adjust the therapy as appropriate or to include other measures, such as more intensive care. This can improve treatment success, patient satisfaction, and quality of life while minimizing the severity of side effects. A more efficient therapy can also reduce costs, for example by eliminating the need for additional treatments. Thus, more successful therapies can relieve the healthcare system as a whole.
These ideas should not just remain on the concept paper. For this reason, Fraunhofer IKS together with Flying Health aims to further develop the ideas and to transfer them into research and development projects. By applying safe intelligence technologies developed by Fraunhofer IKS, it can be shown how the technical and ethical use of artificial intelligence can represent added value for society. In this way, the ideas are intended to be gradually developed into concrete solutions that are relevant and successful on the healthcare market.
This project was funded by the Bavarian State Ministry of Economic Affairs, Regional Development and Energy as part of the project Support for the Thematic Development of the Institute for Cognitive Systems.
