Sci Rep. 2025 Sep 26;15(1):33276. doi: 10.1038/s41598-025-18097-5.
ABSTRACT
Minimizing ionizing radiation is crucial in paediatric imaging due to children’s increased radiation sensitivity, especially at younger ages. To evaluate a CT attenuation-based Auto Prescription protocol for paediatric head imaging, testing whether it provides image quality and radiation dose comparable to age-based protocols. Auto Prescription was implemented on a 256-slice scanner for axial volumetric head CT, adjusting kV and mAs based on attenuation data from scout images. Radiation dose parameters (CTDIvol, SSDE, ED, DLP) and image quality metrics (SNR, CNR, subjective Likert scale from 1-unacceptable to 4-higher than needed) were assessed in 79 consecutive studies using Auto Prescription protocols. These were compared to 68 studies obtained with age-based protocols using non-parametric tests. A total of 147 patients (60 females, mean age 6.7 ± 5.1 years) were included. The auto prescription group included 29 patients aged 0-5, 20 aged 5-10, 25 aged 10-15, and 5 over 15 years; the age-based group included 36, 18, 9, and 5 patients respectively in the same age groups. The Auto Prescription protocol achieved a more balanced radiation dose distribution across age and water-equivalent diameter. The greatest dose reduction was observed in the 0-1 year (48.2%, p < 0.001) and 10-15 year (40.4%, p < 0.001) age groups. While diagnostic image quality was adequate in both settings, it was lower with the auto prescription protocols (mean image quality 3.0 ± 0.2 versus 2.8 ± 0.2; SNR 7.2 ± 1.5 vs. 5.1 ± 1.1; CNR 0.9 ± 0.5 vs. 0.7 ± 0.2; all p < 0.001). All subjective image quality parameters were statistically non-inferior to the age- based protocol (p < 0.05). Attenuation-based Auto Prescription resulted in a more homogeneous and head density adapted radiation dose across paediatric patients, with non-inferior image quality. Dose reduction was a secondary benefit of individualized scan settings based on patient attenuation rather than age alone.
PMID:41006436 | DOI:10.1038/s41598-025-18097-5
