AI Driven Clinical Communication

Motivation

The healthcare ecosystem is a complex network of participants, including patients, doctors, and other stakeholders involved in medical care. In this network, information is shared through established infrastructures by means of electronic health records including radiology reports, discharge letters or lab results. Making sense of this vast amount of information may pose a challenge in time-boxed situations such as a doctor’s appointment: a doctor may not find the time to get a proper overview of a patient’s medical history; a patient may not fully understand its records depending on their health literacy or medical background. Hybrid AI systems integrate data-driven insights with expert knowledge and thus have the potential to contribute to more informed decision making. Embedding these systems into healthcare environments responds to different perspectives as well as expectations.

Healthcare professional perspective: Time constraints in patient care
Project Overview

Studies highlight the consequences of limited time available for patient-doctor communication, making optimal preparation crucial for both. For clinicians, this involves acquiring a thorough understanding of the patient’s medical history (§51(1) ÄrzteG). However, reviewing numerous clinical documents to grasp a patient’s medical
situation is often infeasible due to time limitations. These time constraints may influence the decision-making process, as healthcare professionals must make quick, informed decisions based on incomplete or fragmented data.

Patient perspective: The Impact of Low Health Literacy on Care

Due to varying levels of health literacy in the population and the limited time available for individual consultations with healthcare professionals, much of the clinical information remains unclear. Frequently, uncertainties and fears of those affected are reinforced by misinterpretations or misinformation of health data using unreliable digital sources (social media, Dr. Google, ChatGPT). This has a lasting negative impact on the acceptance of medical treatments, the course of the disease and, as a result, the recovery process, thus leading to an additional burden on the healthcare system.

 

Project Overview
Hybrid AI’s potential in transforming the healthcare sector

AI solutions, particularly those leveraging natural language processing (NLP) with symbolic components, present significant opportunities to enhance access to clear, understandable health information. Healthcare is particularly suited for hybrid AI systems due to the vast amount of structured and unstructured knowledge from ontologies like UMLS and SNOMED CT, and clinical guidelines. Additionally, healthcare requires high levels of transparency, trust, and interpretability in digital systems. By combining data-driven insights with domain knowledge, AI systems can help healthcare professionals make informed decisions by offering well-organized, summarized information. This can improve communication, bridge knowledge gaps, and foster trust and acceptance of AI technologies among both patients and professionals. Integrating AI into existing healthcare platforms, such as public (e.g., ELGA) and private healthcare portals (e.g., x-tention’s eHealth Suite), is essential for expanding accessibility and enhancing digital services. This integration also fosters technological sovereignty by embedding innovative AI-driven solutions into well-established systems.

AID-CC addresses these challenges by improving communication between patients and healthcare professionals through advancements in NLP and hybrid AI, with a focus on German clinical data. By utilizing GDPR-compliant local clinical data in combination with prior knowledge, the project aims to improve health literacy, optimize patient-doctor communication, and streamline information retrieval for healthcare professionals, reducing their workload. To ensure comprehensive compliance and ethical integration, the project not only prioritizes data protection but also adheres to national and EU regulations, such as the Medical Device Regulation and the EU AI Act. An interdisciplinary work package guides the development from planning to full integration within the healthcare ecosystem. These interdisciplinary efforts are
critical for both restoring trust in AI systems, vital for their adoption, and advancing technological sovereignty in healthcare.