Distributed model guided medicine

One of the concerns for providing optimal health care, is the complexity and variety of health care situations and processes. In order to enable intelligent decision making in imaging-based medical diagnosis and therapy, these phenomena may be handled with a model-centric (holistic) approach, also referred to as Model Guided Medicine (MGM). <br/> <br/> Before a trusted application of digital models in health care can be assured, a number of issues need to be addressed, such as finding the right methods and tools for distributed model services, developing multicentre AI models as a base for associated clinical studies, securing human curated transitions in the path from data to models, enhancing model truthfulness through transparency, relevance and explainability as well as providing for a model quality assurance, i.e. model verification, validation and evaluation. <br/> <br/> An IT architecture as presented by a Medical Information and Model Management System (MIMMS) complying with standards relating to AIMS (ISO/IEC 42001) is suggested as a solution concept to the above challenges. In the context of an MGM based on the MIMMS infrastructure and corresponding MIMMS mathematical model categories, attempts are being made in the immediate future to establish a (distributed) Model Quality Assurance Review Centre (MQARC) or an equivalent platform for addressing modelling in the domains of radiology and surgery. Some aspects on how model truthfulness may be handled in such a setting are presented in this paper.