A novel method for EPID transmission dose generation using Monte Carlo simulation and deep learning

Abstract: This study aimed to integrate Monte Carlo(MC) simulation with deep learning(DL)-based denoising techniques to achieve fast and accurate prediction of high-quality electronic portal imaging device (EPID) transmission dose (TD) for patient-specific quality assurance (PSQA). A total of 100 lung cases were used to obtain the noisy EPID TD by the ARCHER MC code under four kinds of particle numbers (1 × 10⁶,1 × 10⁷,1 × 10⁸ and 1 × 10⁹), and the original EPID TD was denoised by the SUNet neural network. The denoised EPID TD was assessed both qualitatively and quantitatively using the structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and gamma passing rate (GPR) with respect to 1 × 109 as a reference. The computation times for both the MC simulation and DL-based denoising were recorded. As the number of particles increased, both the quality of the noisy EPID TD and computation time increased significantly (1 × 10⁶: 1.12 s, 1 × 10⁷: 1.72 s, 1 × 10⁸: 8.62 s, and 1 × 10⁹: 73.89 s). In contrast, the DL-based denoising time remained at 0.13–0.16 s. The denoised EPID TD shows a smoother visual appearance and profile curves, but differences between 1×10⁶ and 1 × 10⁹ still remain. SSIM improves from 0.61 to 0.95 for 1 × 10⁶, 0.70 to 0.96 for 1 × 10⁷, and 0.90 to 0.97 for 1 × 10⁸. PSNR increases by > 20% for 1 × 10⁶ and 1 × 10⁷, and > 10% for 1 × 10⁸. GPR improves from 48.47% to 89.10% for 1×10⁶, 61.04% to 94.35% for 1×10⁷, and 91.88% to 99.55% for 1×10⁸. The method that combines MC simulation with DL-based denoising for EPID TD generation can accelerate TD prediction and maintain high accuracy, offering a promising solution for efficient PSQA.