Integrating AI in Healthcare: Innovations from DMEA 2025

The DMEA 2025 conference has brought to light various innovative products developed by startups aimed at enhancing the integration of artificial intelligence (AI) within healthcare settings, specifically in medical practices and hospitals. A prominent focus has been on intelligent and secure voice processing technologies.

Among the showcased innovations are the AI models Argo and Aureon from IDM gGmbH, which is a nonprofit subsidiary of the University Medical Center Hamburg-Eppendorf (UKE). These AI solutions are designed to assist healthcare professionals with documentation tasks, thereby alleviating their workload. Argo is intended to streamline the creation of medical reports, while Aureon is a text-to-speech model trained specifically on complex medical terminology, making it applicable in various domains including nursing.

The development of these tools involved training a large language model (LLM) using data sourced from UKE, ensuring that the AI can accurately reflect the needs of its users.

During his presentation, the CEO of IDM highlighted the financial and ethical implications of AI systems, pointing out that often, healthcare organizations end up paying for these technologies in two ways: financially and through the provision of data. He emphasized that much of this data and financial investment flows to large tech companies beyond Europe, which does not benefit local healthcare systems. His vision is to develop AI that adheres to European data protection and quality standards, particularly crucial in the medical field. At UKE, a fully digital patient record system has been in operation since 2009, which allows the AI model to access comprehensive patient documentation to generate clinical summaries.

Aureon has been made available to all 15,000 employees at UKE since January of this year, and it has been successfully utilized by nursing staff and administrative personnel. The CEO also announced plans to partially release Aureon as open-source software, aiming to ensure that no healthcare provider in the German-speaking region is deprived of access to speech-to-text technology.

In addition to Aureon, UKE has implemented the transcription tool Orpheus, which has been operational since the beginning of the year.

Another company, Docport GmbH, provided insights into the digital transformation of medical practices, featuring solutions that range from medical technology to practice management systems based on a subscription model. According to one of the founders and a practicing physician at Docport, the control over processed data remains with the medical practices. By employing federated learning techniques, Docport ensures that data does not need to be sent to the AI cloud; instead, AI models are brought to the data within the practices. This approach has enabled the aggregation of millions of data points annually, and the combination of data analytics with optimized processes is viewed as potentially lifesaving.