Pitfalls and Practical Protocol Optimization for Low-Pitch Spiral 4DCT Mode: Assessment of Static Image Quality In Terms of Noise, Artifacts, and Slice Sensitivity Profile
Abstract
Purpose
To assess how low-pitch spiral 4DCT impacts image quality, particularly noise, artifacts, and slice sensitivity profile (SSP), to help overcome the limitations of this technique.
Methods
Two CT scanners, the Siemens Biograph PET/CT (S-type scanner) and the Philips Brilliance Big Bore CT (P-type scanner), were used with low-pitch spiral 4DCT protocols on phantoms of various water-equivalent diameters (WED: 27, 30, 33 cm). Noise was measured using a global noise index (GNI), and artifacts were assessed qualitatively. Z-resolution was compared across axial, spiral, and low-pitch 4DCT modes with an SSP phantom. Correlations between GNI and scan parameters were analyzed with Spearman’s rank test.
Results
In the P-type scanner, GNI correlated strongly with mA (-0.836) and pitch (-0.793), moderately with WED (0.453), and weakly with CTDIvol (-0.192) and gantry cycle (0.245). The GNI in the S-type scanner correlated strongly with CTDIvol (-0.618), and moderately with mA (-0.536) and WED (0.442). The cutoff GNI ≤ 20 in P-type is pitch ≥ 0.06 and mAs ≥ 800, and is an Automatic Exposure Control (AEC) level of QRF ≥ 70 in S-type. A prominent ring artifact was consistently observed at mA < 100 for WED=33 cm in P-type, but S-type showed no artifacts under similar low mA conditions. P-type demonstrated sharper SSP in axial mode, whereas S-type better preserved SSP shape in spiral mode. For both scanners, the full width at half maximum (FWHM) of the SSP increased slightly in low-pitch 4DCT mode but remained within 0.3 mm of the nominal slice thickness.
Conclusion
The methods enable the identification of mA–pitch–AEC operating ranges that cause artifacts. The results define scanner-specific thresholds at which low-pitch operation introduces noise or ring artifact, providing practical boundaries for clinically reliable acquisition. The effects of low mA settings should be evaluated for new scanners using this approach.