Responsible Imaging AI and Clinical Decision Support in Radiology and Radiotherapy

Medical imaging is central to modern radiotherapy, enabling tumour delineation, treatment planning, image guidance, adaptive radiotherapy, treatment verification, and response assessment. Advances in CT, CBCT, MRI, PET, quantitative imaging, machine learning, deep learning, foundation models, agentic AI, and large language models are reshaping these workflows through automated analysis, reconstruction, multimodal interpretation, and clinical decision support. However, important challenges remain concerning generalisability, uncertainty, bias, interpretability, data privacy, cybersecurity, regulatory compliance, and the safe integration of AI into clinical practice. Recent studies have demonstrated promising improvements in segmentation, registration, image enhancement, radiomics, treatment adaptation, and workflow efficiency, but evidence of clinical benefit remains variable across institutions, scanners, patient populations, and disease sites. Responsible implementation therefore requires rigorous external validation, transparent evaluation, meaningful human oversight, and continuous monitoring after deployment. Further investigation is needed to establish robust standards for trustworthy imaging AI and to clarify how these systems can support, rather than replace, expert judgement in radiology and radiotherapy.This Research Topic aims to advance the development, validation, and clinical translation of responsible imaging AI for radiology and radiotherapy. It will address how AI can improve the accuracy, efficiency, consistency, and personalisation of imaging-driven cancer care while maintaining safety, fairness, transparency, privacy, and accountability. Contributions may examine methodological innovations, clinical utility, human–AI collaboration, implementation strategies, and governance frameworks. Particular emphasis will be placed on robust validation, uncertainty quantification, explainability, multimodal reasoning, workflow integration, and evidence demonstrating benefit in realistic clinical settings.To gather further insights in responsible imaging AI for radiology and radiotherapy, we welcome articles addressing, but not limited to, the following themes:• AI for medical image segmentation, target delineation, organ-at-risk contouring, uncertainty estimation, interactive segmentation, and foundation models.• AI-based image registration, deformable registration, multimodal image fusion, robustness evaluation, and quality assessment.• AI for image reconstruction, denoising, super-resolution, synthetic imaging, enhancement, and harmonisation.• Imaging AI for image-guided radiotherapy, CBCT analysis, motion management, treatment verification, anatomical change detection, and adaptive image guidance.• AI-enabled adaptive radiotherapy, automated replanning, dose accumulation, image-based treatment adaptation, and patient selection.• MRI-guided radiotherapy, synthetic CT generation, quantitative MRI, multimodal imaging, and imaging protocol optimisation.• Radiomics and quantitative imaging biomarkers for treatment response, prognosis, toxicity assessment, and personalised radiotherapy.• Large language models, vision-language models, conversational AI, and agentic AI for imaging interpretation, reporting, workflow orchestration, clinical decision support, and human–AI collaboration.• Imaging quality assurance, protocol standardisation, scanner harmonisation, and automated cross-institutional quality assessment.• Explainable, trustworthy, and responsible imaging AI, including uncertainty quantification, bias mitigation, fairness, privacy-preserving AI, cybersecurity, governance, and regulation.• Clinical validation, multicentre evaluation, reproducibility, external validation, federated learning, and deployment across diverse imaging environments.• Human–AI interaction, usability, clinician acceptance, reader studies, observer variability, and workflow integration.• Data standards, interoperability, collaborative imaging infrastructures, and clinical implementation.• Post-deployment monitoring, model drift detection, quality management, and continuous evaluation in radiology and radiotherapy.We welcome Original Research, Technical Reports, Systematic Reviews, Perspectives, and Brief Research Reports. Topic editors: James C. L. Chow, Kay Li, Haya Aldossary Source: https://www.frontiersin.org/research-topics/84729/responsible-imaging-ai-and-clinical-decision-support-in-radiology-and-radiotherapy