Advanced Dose Calculation Strategies for Clinical Linear Accelerators: A Systematic Review

This systematic review evaluates the current evidence on Monte Carlo (MC)-based dose calculation strategies for clinical linear accelerators (LINACs) in radiotherapy. The review focuses on developments from 2010 to 2025 that enhance dose accuracy and workflow efficiency, particularly through GPU acceleration and artificial intelligence (AI). Using an adapted PICO framework for technical/methodological reviews, the review maps MC implementations across major LINAC platforms (Varian, Elekta, Siemens), assesses dosimetric accuracy improvements over conventional algorithms, evaluates computational acceleration strategies, and identifies clinical scenarios where MC provides meaningful accuracy gains exceeding 3-5% dose difference thresholds. Data synthesis follows the Synthesis Without Meta-analysis (SWiM) guidelines due to the heterogeneity of study designs and outcome metrics. A custom quality assessment framework informed by the AAPM TG-268 RECORDS checklist, AHRQ modeling/simulation guidance, and JBI critical appraisal tools is used to evaluate included studies across four domains: methodological rigor, validation completeness, clinical relevance, and uncertainty analysis. The review is reported following PRISMA 2020 guidelines.