Silver Filter with Deep Learning Reconstruction Lowers Abdominopelvic CT Dose by 36%

Noncontrast abdominopelvic CT with silver filter + deep learning reconstruction cuts radiation dose by 36% (p≤0.05) without image quality loss; HU values significantly lower for urine (38%) and subcutaneous fat, relevant for cyst/adrenal evaluation.

• Radiation dose (CTDIvol, DLP) was 36% lower with the silver filter and deep learning reconstruction compared with conventional filtration (p≤0.05).
• Only urine (38% decrease) and subcutaneous fatty tissue showed statistically significant HU reduction; other soft tissues had non-significant changes.
• The altered Hounsfield units could affect evaluation of indeterminate cysts or adrenal masses on unenhanced scans.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.