S2995
Invited Speaker
ESTRO 2026
adaptive and data-informed radiotherapy. This presentation outlines how AI-enabled tools can be embedded across the radiotherapy pathway to improve efficiency, consistency and responsiveness while maintaining patient safety and clinical accountability. In simulation and planning, automated segmentation and contouring support can reduce time spent on routine delineation and enable RTTs to focus on image interpretation, uncertainty management, protocol adherence, peer review and escalation. In treatment delivery, AI-assisted image guidance and adaptive workflows can streamline plan selection, online adaptation and structured checks, positioning RTTs to coordinate adaptive pathways, triage complexity, and standardise decision points in collaboration with radiation oncologists and medical physicists. Clinical decision support can assist workload prioritisation and outlier detection; however, advanced practice RTT roles are critical for local commissioning/validation, performance monitoring, and governance processes that define responsibilities, thresholds and documentation. Automation of documentation and structured data capture can reduce administrative burden and strengthen traceability, while scheduling optimisation can better match machine capacity, acuity and staffing. Education and credentialing must evolve: RTTs require competencies in model limitations, bias, data quality, human factors, and ongoing QA to critically appraise outputs rather than accept them unchallenged. Finally, AI-enabled analytics can support incident learning and continuous improvement by translating workflow data into actionable insights. With human-centred design and robust change management, RTT-led adoption of AI and automation can expand advanced practice, support safer adaptive radiotherapy, and deliver equitable, high-quality care at scale. 5302 The quest for predictive biomarkers: Is the historical one-size-fits-all-approach finally over? Ingeborg Tinhofer Department of Radiooncology and Radiotherapy, Charité - Universitätsmedizin Berlin, Berlin, Germany Resectable head and neck squamous cell carcinoma (HNSCC) continues to be managed predominantly according to stage- and pathology-based risk stratification. Although this approach has provided an effective framework for surgical and postoperative decision-making, it does not adequately explain the substantial heterogeneity in outcomes observed among patients with apparently similar disease. Established clinicopathologic factors, including margin status, extranodal extension, nodal burden, HPV status, and eligibility for radiotherapy or cisplatin,
evidence base, and practical considerations for SFRT implementation. Key prospective and real-world studies, including DART, FAST-METS, and SFRT-1000, demonstrate feasibility, safety, and reduced treatment burden, with acceptable dosimetric accuracy in palliative settings. Both adaptive and non-adaptive workflows are examined, including diagnostic image– based pre-planning, atlas-based approaches, and on- table adaptation using cone-beam CT (CBCT) or magnetic resonance (MR) imaging. Consensus recommendations derived from a modified Delphi process are presented, defining requirements for safe and scalable implementation across domains including patient selection, imaging quality and recency ( ≤ 30 days), workflow standardization, and quality assurance (QA). Practical implementation strategies emphasize aligning technical complexity with anatomical risk, prioritizing rigid targets (e.g., spine and pelvis), and embedding clear fallback pathways. Comparative considerations between adaptive and IGRT-based approaches are addressed. While adaptive platforms enable real-time optimization, limited availability reinforces the role of IGRT-based SFRT as a pragmatic and scalable solution that supports equitable access without compromising safety. Emerging applications in stereotactic ablative radiotherapy (SABR) are discussed, alongside ongoing challenges in motion management and complex anatomy. Conclusion SFRT represents a clinically mature and scalable model of radiotherapy delivery. With appropriate safeguards, both adaptive and non-adaptive workflows can be integrated into routine practice today. As evidence, consensus, and technology continue to evolve, SFRT provides a foundation for more responsive, efficient, and accessible radiotherapy pathways. 5299 AI and automation in clinical workflows: Shift from manual tasks to evaluative decision-making by RTTs Nigel J Anderson Radiation Oncology, Icon Group, QUEENSLAND, Australia Artificial intelligence (AI) and automation are rapidly reshaping radiation therapist (RTT) clinical workflows, shifting practice from predominantly manual, repetitive tasks toward advanced, evaluative decision- making and leadership in service delivery. For the RTT community, AI integration is not only a technology change but a scope-of-practice opportunity: enabling RTTs to lead protocol-driven implementation, quality assurance (QA), education, and safe adoption of
Made with FlippingBook - Share PDF online