Impact of Knowledge-Based Iterative Model Reconstruction on Image Quality and Hemodynamic Parameters in Dynamic Myocardial Computed Tomography Perfusion Using Low-Tube-Voltage Scan: A Feasibility Study.

Abstract

Knowledge-based iterative model reconstruction (IMR) yields diagnostically acceptable image quality in low-dose static computed tomography (CT). We aimed to evaluate the feasibility of IMR in dynamic myocardial computed tomography perfusion (CTP). We enrolled 24 patients who underwent stress dynamic CTP using a 256-slice CT. Images were reconstructed using filtered back projection (FBP), hybrid IR, and IMR. Image quality and hemodynamic parameters were compared among three algorithms. Qualitative image quality and contrast-to-noise ratio were significantly higher by IMR than by FBP or hybrid IR (visual score: 4.1 vs. 3.0 and 3.5; contrast-to-noise ratio: 12.4 vs. 6.6 and 8.4; P < 0.05). No significant difference was observed among algorithms in CTP-derived myocardial blood flow (1.68 vs. 1.73 and 1.70 mL/g/min). The use of knowledge-based iterative model reconstruction improves image quality without altering hemodynamic parameters in low-dose dynamic CTP, compared with FBP or hybrid IR.