Digital Twins and Knowledge-Based AI for Nuclear Medicine: Standards, Ontologies and Clinical Translation

Artificial intelligence (AI) is poised to transform nuclear medicine along multiple, interconnected directions. On the translational side, AI can accelerate and optimise clinical trial design, treatment planning, and response prediction in ways that parallel its breakthrough impact on protein-folding prediction and molecular design. In nuclear medicine, this evolution may culminate in the development of digital twins that faithfully replicate human physiology and disease, beginning with preclinical animal models and extending to individual patients. At the same time, advancing the foundations of knowledge requires rigorous ontological representations of nuclear medicine concepts, procedures, radiobiological models, and data flows. Such ontologies can provide shared, machine-readable frameworks that support AI development, ensure semantic consistency, and act as transparent “reference maps” for validating digital twins and other AI-based tools.This Research Topic aims to address a central challenge: how to integrate AI, digital twins, and knowledge-based systems into nuclear medicine in a way that is scientifically robust, clinically useful, ethically sound, and globally interoperable. Despite rapid progress, several key gaps remain. Digital twins are often developed using heterogeneous data, non-standardised protocols, and radiobiological models that are not entirely validated for AI-driven applications. Ontologies for nuclear medicine are still fragmented or incomplete, limiting the ability to share, reuse, and compare data and models across centres and studies. In addition, regulatory, ethical, and safety frameworks are struggling to keep pace with technical innovation, especially when patient-specific simulations and decision-support systems are involved. The goal of this Research Topic is to bring together multidisciplinary contributions that propose methods, frameworks, and case studies for building reliable digital twins, developing and using ontologies, and defining best practices, guidelines, and evaluation criteria for AI in nuclear medicine.We welcome original research, reviews, methods papers, perspectives, and case reports that address AI-enabled nuclear medicine with particular emphasis on digital twins and knowledge representation. Topics of interest include, but are not limited to:• Design and validation of digital twins as human model replicas for nuclear medicine• Development of guidelines, harmonised rules, and unified software/platform ecosystems• Integration of radiobiological and dosimetric models into AI-driven pipelines• International cooperation frameworks and multi-centre initiatives• Ontology development and semantic standards for nuclear medicine• Evaluation of AI-based techniques, accuracy, robustness, and generalisability• Case studies, proofs of concept, and safety/uncertainty analyses• Ethical, legal, and societal aspects of AI and digital twins in nuclear medicine• Challenges related to data banks, curated datasets, and data-sharing policiesDr. Vijay Shah is employed by Siemens Medical Solutions USA, Inc. All other Topic Editors declare no competing interests with regards to the Research Topic subject. Submissions that bridge basic, translational, and clinical perspectives are particularly encouraged. Topic editors: Alessandro Bartoloni, Vijay P Shah, Lidia Strigari Source: https://www.frontiersin.org/research-topics/76640/digital-twins-and-knowledge-based-ai-for-nuclear-medicine-standards-ontologies-and-clinical-translation