Application of Tin Spectral Filtration for CT Colonography: Impact on Image Quality and Radiation Dose.

RATIONALE AND

OBJECTIVES

The purpose of this study was to evaluate image quality and radiation dose when using tin spectral filtration for computed tomography colonography (CTC) screening.

MATERIALS AND

METHODS

A phantom wrapped in fat-attenuation rings simulated five patient sizes. Noise-dose relationships were acquired by scanning the phantom using conventional 120-kV and tin-filtered 100-kV (100Sn) and 150-kV (150 Sn) protocols at 1 mGy volume CT dose index (CTDIvol) increments. These protocols were applied in 100 patients undergoing CTC in supine and decubitus positions, with each position randomly assigned to standard or tin-filtered protocols. Seven readers performed blinded side-by-side comparisons, scoring colon segments and extracolonic findings on a 5-point Likert scale. Proportions of cases where tin-filtered images were comparable or better were calculated. CTDIvol, dose-length product (DLP), and effective dose, and organ doses were assessed.

RESULTS

For colon segments, the proportions of tin-filtered cases with comparable or better image quality were 0.940 to 0.981 for endoluminal views and 0.927 to 0.987 for axial images. For extracolonic findings, the proportions of tin-filtered cases with comparable or better image quality were 0.722 for nephrolithiasis, 0.846 for lymphadenopathy, 0.770 for liver lesions, and 0.783 for renal lesions. There were no cholelithiasis cases. Tin-filtered scans reduced CTDIvol, DLP, and effective dose by 37%, 38%, and 29%, respectively. Organ dose reductions were 30% for the colon, 29% for the liver, 28% for the stomach, and 34% for the skin.

CONCLUSION

Optimized tin spectral filtration enables significant radiation dose reduction in CTC without compromising image quality.