Dual-focal-spot single-detector CT: A simulation study to assess cone-beam artifacts and noise

Ultra-wide coverage CT (> 128 detector rows) makes it possible to image a heart or brain in a single rotation but are associated with large cone angles, which can severely degrade the image quality.

This study evaluate the image quality and artifact levels of a dual-focal-spot single-detector (DFSSD) CT geometry designed to achieve 140 mm z-axis coverage, through a simulation study.

The DFSSD CT system employs two x-ray focal spots spaced 90 mm apart along the z-axis and a 100 mm CT detector. It provides 140 mm z-axis coverage at the iso-center. A research CT simulation package (CatSim) was employed throughout this study. To compare the image quality, three system geometries were simulated: (A) single focal spot with 40 mm detector (VCT40), (B) single focal spot with 140 mm detector (VCT140), and (C) the DFSSD CT system. A simulated helical body phantom (45 cm x 30 cm x 16 cm) containing various bony structures was used to assess cone-beam artifacts and noise uniformity across the edge, central, and intermediate slices under different acquisition modes: axial half scan, axial full scan, and helical scan at various helical pitch (0.5, 0.75, 1.0). All images were reconstructed using a cone-beam filtered back-projection algorithm with 3D cone-angle-dependent pixel-wise weighting.

Under helical scan conditions (pitch 0.5 to 1.0), all geometries demonstrated similar levels of cone-beam artifacts and noises. Slightly increased artifacts observed on some slices in VCT140 and DFSSD CT, but overall image quality are acceptable. However, under axial half scan condition, the VCT140 geometry exhibited significantly worse artifacts and image uniformity in intermediate and edge slices compared to VCT40, whereas DFSSD CT showed artifact and image uniformity comparable to VCT40.

Simulation results indicate that the DFSSD CT geometry can achieve 140 mm z-axis coverage while maintaining image quality similar to the VCT40 system.