Impact of 1024-Matrix Size on Perforating Artery Visualisation in Cerebral Computed Tomography Angiography Using a...

Computed tomography angiography (CTA) is essential for preoperative evaluation of intracranial vessels; however, the visualisation of small perforating arteries remains challenging with conventional systems. This study investigated whether increasing the reconstruction matrix size from 512 to 1024 pixels improves the visualisation of perforating arteries in cerebral CTA using a standard 64-slice computed tomography (CT) scanner.

The study comprised phantom and clinical components. Physical image properties were assessed through Task Transfer Function (TTF) and Noise Power Spectrum (NPS) analyses. In the clinical phase, 41 patients underwent cerebral CTA with both 512- and 1024-matrix reconstructions. Quantitative analysis was performed to determine the peak CT values of the anterior choroidal artery and standard deviation in the basal cistern. Two independent observers (CT radiographers) evaluated the visualisation of anterior choroidal and posterior thalamoperforating arteries using a 5-point scale.

Physical metrics showed minimal differences between matrix sizes, with 5% TTF values of 0.862 and 0.867 cycles/mm for 512 and 1024 pixels, respectively. The anterior choroidal artery exhibited higher peak CT values with 1024 pixels (139.31 ± 27.02 HU vs. 136.85 ± 26.69 HU, p < 0.001). The qualitative assessment revealed significant improvements in vessel visualisation with 1024 pixels, particularly for the posterior thalamoperforating artery, where high-quality visualisation increased from 17.07% to 39.03%.

Although increasing the matrix size from 512 to 1024 pixels did not significantly affect the physical image quality metrics, it enhanced the visual detection of small perforating arteries through enhanced sampling density. This optimization technique offers a practical approach for enhancing preoperative vessel evaluation using existing CT technology without requiring specialised equipment.