A comparative study of MTF methods for evaluating spatial resolution in pediatric thoracic phantom CT imaging: effects of acquisition pitch and slice thickness

Purpose. This study developed a pediatric thoracic phantom equipped with a spatial resolution module and caliper tool for evaluating angular and distance accuracies. The effects of pitch and slice thickness on the organ dose and their influence on the Point-, Edge-, and z-modulation transfer functions were analyzed to optimize protocols based on the AAPM Pediatric Chest Routine 2017. Methods. The phantom was fabricated using TCIA patient DICOM data. Eight protocols were tested: P1–P4 (slice thickness 1.25 mm; pitch 0.516, 0.969, 1.375, 1.531) and P5–P8 (pitch 1.375; slice thickness 0.625, 1.25, 2.5, 3.75 mm). Scans were performed at 120 kV, 70 mA, 250 mm FOV, 50 ms rotation, and ASiR-V 50%. The organ doses for the lungs, soft tissue, spine, and heart were measured using Gafchromic LD-V1 film. The image quality was evaluated using image noise, in-plane and z-axis MTF, and caliper-based accuracy tests. Results. CTDIvol decreased with increasing pitch, and P2 (2.94 mGy) met the standard CTDIvol range. Pitch had a minimal effect on the MTF (< 5%), whereas slice thickness caused up to 35% variation. Edge-MTF (p = 0.81; CR = 1.65 for pitch and p = 0.001; CR = 0.58 for slice thickness) showed superior in-plane resolution and acceptable repeatability. z-MTF (p = 0.14; CR = 0.03 for pitch and p = 0.001; CR = 0.04 for slice thickness) demonstrated highly reproducible longitudinal resolution. The noise ranged from 10 to 18 HU, with P2 exhibiting moderate noise (∼16 HU), supporting reliable MTF assessment. Caliper tests revealed an angular deviation of 0.59° and a distance error of 0.32%. Conclusion. Protocol P2 (pitch 0.969, slice thickness 1.25 mm) offers an optimal balance of dose, image quality, and repeatability, achieving edge-MTF of 1.63 ± 0.52 mm⁻¹ and z-MTF of 0.60 ± 0.02 mm⁻¹, ensuring robust spatial resolution for pediatric CT imaging.