儿科光子计数探测器CT辐射剂量与客观图像质量系统综述/meta分析

BACKGROUND: Photon-counting detector computed tomography (PCD-CT) offers potential advantages over conventional energy-integrating detector computed tomography (EID-CT) in pediatric imaging, but the available evidence has not yet been systematically synthesized.

OBJECTIVE: To compare radiation dose, image quality based on objective metrics, and iodinated contrast dose between PCD-CT and EID-CT in pediatric patients.

MATERIALS AND METHODS: A systematic search of PubMed, Scopus, and Cochrane CENTRAL through March 15, 2026, identified comparative studies of PCD-CT versus EID-CT in children 17 years of age or younger. Radiation dose outcomes, including volume computed tomography dose index (CTDIvol), dose-length product (DLP), and size-specific dose estimate (SSDE), were pooled as mean differences (MDs), and objective image quality outcomes, including signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), were pooled as standardized mean differences (SMDs; Hedges' g) using random-effects models. Prediction intervals were calculated for each pooled outcome. Subgroup analyses were performed by body region, and exploratory meta-regression with body region as a moderator was performed.

RESULTS: Twelve comparative studies comprising 1,545 total study-arm observations or examinations were included in the meta-analysis, and 6 additional studies were included in the narrative synthesis. Compared with EID-CT, PCD-CT showed significantly lower CTDIvol (MD=-0.90 mGy; 95% confidence interval [CI]: -1.74 to -0.06; P=0.036; 95% prediction interval [PI]: -4.06 to 2.26) and significantly lower SSDE (MD=-1.35 mGy; 95% CI: -2.32 to -0.38; P=0.006; PI: -4.63 to 1.92). DLP showed a non-significant reduction (MD=-21.25 mGy·cm; 95% CI: -47.29 to 4.79; P=0.110; PI: -116.58 to 74.08). PCD-CT also demonstrated significantly higher SNR (SMD=0.42; 95% CI: 0.05 to 0.78; P=0.024; PI: -0.94 to 1.77) and a small significant increase in CNR (SMD=0.34; 95% CI: 0.02 to 0.66; P=0.037; PI: -0.66 to 1.34). All overall 95% prediction intervals crossed zero. The head subgroup showed the largest objective SNR difference (k=2; SMD=1.34; 95% CI: 1.12 to 1.56). Two studies reported 25-50% reductions in contrast dose while maintaining image quality based on objective metrics.

CONCLUSION: Compared with EID-CT, PCD-CT was associated with significantly lower CTDIvol and SSDE, and with significantly higher SNR and CNR; the reduction in DLP did not reach statistical significance. All overall 95% prediction intervals crossed zero, indicating that the magnitude and direction of benefit may vary across protocols and body regions. The most consistent dose reduction was observed in abdominopelvic imaging, and the largest objective SNR difference was observed in head computed tomography, although the clinical relevance of SNR/CNR differences remains to be established.