Iterative Model Reconstruction (IMR) improves image quality and diagnostic performance in prospective ECG-Gated coronary CT Angiography (CTCA)

Abstract

Background: IMR is a new generation iterative reconstruction method only recently introduced in Australia. The aim of this study was to determine the impact of IMR on image quality and diagnostic performance utility in subjects presenting to our suburban hospital for routine CTCA. Reduction in image noise has the potential to allow lower radiation dose scans, through reducing kVp and mAs, without compromising image quality. Methods: Twenty-five patients underwent a prospective ECG gated CTCA, at tube voltages between 100-120kVp on a Phillips 256-slice CT scanner. Subject weight, heart rate and radiation dose using Dose-Length-Product (DLP) were recorded. Each CTCA was reconstructed with iDose and IMR. Objective image noise in the proximal ascending aorta (PAo) and left ventricular cavity (LVC) were measured. An experienced interpreter compared both reconstruction methods for non-diagnostic segments. Statistical significance was assessed using a paired T test. Results: The mean heart rate was 59.9bpm, mean weight 97.2 kg, and mean DLP was 186mGy*cm. Although 10 subjects weighed more than 100 kg, only one subject was scanned at 120kVp, all others had 100kVp scans. Compared with iDose, image noise with IMR was significantly lower in both the PAo (IMR noise 22.8+/-6.5 HU versus iDose 40.2+/-16.1) (P<0.01) and LVC (IMR noise 23.6 +/-5.8 HU versus iDose 47.6+/-14.8 HU) (P<0.01). Overall, there were fewer non-diagnostic segments with IMR (4) compared with iDose (16), (P<0.01). Conclusions: IMR improves image quality and diagnostic performance in prospective ECG-gated CTCA and therefore has the potential to allow lower radiation dose scans without compromising image quality.