Sung Bin Park MD, Presenter: Nothing to Disclose

Seong Hoon Choi, Abstract Co-Author: Nothing to Disclose

Jong Beum Lee, Abstract Co-Author: Nothing to Disclose

Hyun Jeong Park, Abstract Co-Author: Nothing to Disclose

The purpose of this study was to evaluate the efficacy of knowledge-based iterative model reconstruction (IMR) algorithm for reducing image noise in ultra low-dose non-enhanced CT (ULDCT) and the diagnostic performance of ULDCT for urolithiasis.

103 patients diagnosed with urinary stones (n=284) using both a standard-dose non-enhanced CT (SDCT, 120 kV and 150mAs) and an ultra low-dose non-enhanced CT (100kV and 20mAs) at two institutions were enrolled in the study. SDCT images were reconstructed with filtered back projection (FBP), and ULDCT images were reconstructed with FBP, hybrid iterative reconstruction (iDose4 level 5), and the IMR (body soft tissue level 3) algorithm. Interpretations of the two scans were performed prospectively with respect to radiation dose, objective image noise, and subjective image assessment (image quality, noise, diagnostic confidence). With SDCT-FBP as the reference standard, diagnostic performance and inter-observer agreement of ULDCT-IMR were assessed between two reviewers.

The average effective dose of SDCT and ULDCT was 8.31 mSv, and 0.68 mSv, respectively, and the average radiation dose reduction rate was 91.82% (p<0.01). Objective image noise was lower in ULDCT-IMR (p<0.01) than SDCT-FBP as well as ULDCT-FBP and ULDCT-iDOSE. The subjective assessment in ULDCT-IMR was comparable to that of SDCT-FBP, although SDCT-FBP was still superior statistically. Among 284 urinary stones detected by SDCT-FBP, 229 (80.6%) were detected by ULDCT-IMR, in which detection percentage were 66/69 (95.7%) for ureter stones and 155/207 (74.9%) for kidney stones. Non-detectable stones were 3 mm or less in size, which are clinically insignificant. Inter-observer agreement of ULDCT-IMR between the two reviewers in the diagnosis of stones was high with kappa values (kappa = 0.82, excellent).

ULDCT-IMR provided a significant reduction in radiation dose while maintaining diagnostic performance and image quality comparable to that of SDCT-FBP for diagnosing urinary stones.

Patients with urolithiasis can be evaluated with ultra low-dose non-enhanced CT using knowledge-based iterative model reconstruction algorithm at a substantially reduced radiation dose, thereby minimizing risks to patient from radiation exposure while providing the clinically relevant diagnostic benefits.

Park, S, Choi, S, Lee, J, Park, H, Knowledge-based Iterative Model Reconstruction Algorithm (IMR) for Evaluation of Urolithiasis: With Respect to Radiation Dose Reduction, Image Quality and Diagnostic Performance.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14016464.html