Artificial intelligence in gamma knife radiosurgery: applications in treatment planning, response prediction, and toxicity assessment.
Authors
Affiliations (5)
Affiliations (5)
- Temerty Faculty of Medicine, University of Toronto, Toronto, ON, Canada.
- Department of Radiation Oncology, University of Toronto, Toronto, ON, Canada.
- Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, ON, M4N 3M5, Canada.
- Department of Radiation Oncology, University of Toronto, Toronto, ON, Canada. [email protected].
- Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Toronto, ON, M4N 3M5, Canada. [email protected].
Abstract
Artificial intelligence (AI) is increasingly being developed across the Gamma Knife radiosurgery (GKRS) workflow, but its current applications and readiness for clinical integration remain underutilized. This narrative review examines the emerging role of AI in GKRS, with emphasis on treatment planning, outcome prediction, and post-treatment assessment. A structured search of PubMed/MEDLINE, Embase, and Scopus was conducted through September 20, 2026 to identify studies applying learning-based AI approaches specifically to GKRS. Eligible studies evaluated AI models using GKRS clinical, imaging, or treatment-planning data, while non-GKRS and non-learning-based studies were excluded. Included studies were synthesized according to major stages of the GKRS workflow. AI applications were identified across tumor segmentation, isocenter selection and treatment planning, prescription and dose-distribution prediction, synthetic computed tomography generation, and plan-quality assessment. Post-treatment applications include longitudinal tumor monitoring and prediction of response, functional outcomes, and toxicity including peritumoral edema and radiation necrosis. Several models demonstrate promising performance, but the evidence remains predominantly retrospective and task-specific, with frequent reliance on single-center or small datasets and limited external or prospective validation. AI has the potential to improve the consistency, efficiency, and personalization of GKRS by supporting specific technical and clinical decisions. Further multicenter validation, prospective evaluation, workflow integration, and demonstration of meaningful clinical benefit are required before routine adoption. AI currently remains a decision-support tool rather than a replacement for radiosurgical expertise.