Digital Twins in eHealth – : Prospects and Challenges Focussing on Information Management

A new approach for managing the knowledge of artificial intelligence based, learning ehealth systems via digital twins will be presented. With different subtypes of such twins: personal, group and system digital twins, a number of decisive targets can be achieved in addition to the ubiquitous knowledge aggregation and organization by such twins. These targets start with allowing to forget respectively unlearn person-related data as mandated by new EU-GDPR. The targets include also enhanced transparency of the system for delivering unbiased conclusions within the limitations of the designated target group with respect to the learned training data. Finally, a self-monitoring of the assessed efficacy of the system in the presence of machine learning will be tackled, including also the reapplication of prior conclusions (diagnoses) in the light of the additional knowledge learned.