Optimizing Image Quality in Carbon Nanotube-Based Rectilinear Digital Tomosynthesis System via Projection Number Variation

External Beam Radiation Therapy (EBRT) is widely used in cancer treatment, requiring precise patient alignment for optimal outcomes. Traditional Cone-Beam Computed Tomography (CBCT) in Image-Guided Radiotherapy (IGRT) faces challenges like high radiation exposure and lengthy acquisition times. This study explores a rectilinear Digital Tomosynthesis System (DTS) with carbon nanotube (CNT)-based X-ray sources as an alternative for EBRT. By varying projection numbers, we assessed metrics including Structural Similarity Index Measure (SSIM), Signal-to-Noise Ratio (SNR), and Artifact Spread Function (ASF). Results indicated that increasing projection numbers enhanced SSIM, SNR, and artifact suppression, balancing image quality and computational efficiency. Integrating GPU acceleration further optimized processing speed, supporting real-time patient positioning in radiotherapy without conventional on-board imaging.