SU‐GG‐I‐12: Effect of Geometric Unsharpness On the Reconstructed Image in Region of Interest (ROI) μCT

Abstract

Purpose: To demonstrate the effects of variation of focal‐spot size and magnification on the spatial resolution of reconstructed images of a micro‐computed tomography (μCT) system which is attached to a standard angiographic C‐arm gantry to enable Region‐of‐Interest cone‐beam CT (ROI‐CBCT). Method and Materials: High‐resolution ROI projection data of a vascular phantom were acquired using a new high‐sensitivity, microangiographic fluoroscope (HSMAF) detector (35 μm pixels), which was attached to the C‐arm gantry and able to be positioned in front of a standard full field‐of‐view, low‐resolution commercial flat‐panel detector (FPD) (194 μm pixels). The HSMAF consists of a CsI phosphor viewed by a 4‐cm diameter light image‐intensifier with large variable dynamic range whose output is coupled via a fiber‐optic taper to a CCD camera. The test objects in the vascular phantom were a stent (100 micron struts) inside of a catheter in a cylindrical water bath. The phantom was placed on a portable test platform (PTP) enabling CBCT image acquisition by the HSMAF every 1°. Six μCT runs were performed using two focal‐spot sizes (0.3 and 0.6 mm) and three magnification factors (1.15, 1.29, and 1.48). Profiles were extracted from the reconstructed struts, and the full width half‐maximum (FWHM) were measured. Results: The reconstructed data show that using the optimal configuration (smallest magnification with small focal spot) compared to the worst configuration (largest magnification and the large focal spot) resulted in a 47% reduction in the FWHM in the object plane (175 μm versus 375 μm). Conclusion: Micro‐CBCT can provide more accurate visualization of fine device features; however, geometric unsharpness and/or large focal spots can substantially degrade resolution reducing the quality of the μCBCT reconstructions.(Research sponsored by: NIH Grants R01‐NS43924, R01‐EB002873, Toshiba Medical Systems Corporation)