ELSA-AidArtificial intelligence in diagnostics: ethical, professional and social aspects

ELSA-AID Logo Copyright: © Helen Heinrichs

Artificial intelligence (AI) holds promise to improve health care delivery and will be an important
enabler of precision medicine. For diagnostics in particular, the potential of machine learning to
automatically evaluate and merge different diagnostic data to benefit the patient is tremendous. At
the same time, there are pressing challenges inherent to implementing AI-aided diagnosis: Different medical specialties need to cooperate more intensely, and they need to integrate the AI systems into their routines in clinical practice. This will reshape relationships in the medical realm, a)
incorporating AI into physician-physician and b) into physician-patient relationships, posing new
questions about responsibility, and c) reframing what it means to trust a diagnostic decision. Despite tremendous advances in developing AI-aided precision medicine, the ethical, social, and professional consequences, crucial for implementation in routine medical care, are hardly investigated. This project will approach those complex interactions with empirical and normative studies and
suggest measures to realize benefits for individual and society to the fullest. We will evaluate how
interfaces between machine and human in AI-aided diagnosis can ideally be composed. We will
implement demonstrators to provide a basis for an in-depth discussion of transparency and traceability for self-learning systems for medical decisions, for dissemination of possible implementation
scenarios, and ultimately to help define the role of AI-aided diagnostics within digital medicine.

Further information can be found on the Website of the BMBF or of our project partner at the institute of Applied Ethics or at our third research Partner the Fraunhofer MEVIS .