Effect of Iterative Reconstruction Algorithm in Image Space on Image Quality In Abdominal-Pelvic CT wWith Contrast

 Iterative reconstruction algorithms have been suggested to reduce the dose in CT scans, but their effect on image quality needs more investigation. This study evaluates the impact of the Iterative Reconstruction in Image Space (IRIS) algorithm on image quality in contrast-enhanced abdominal-pelvic CT.

Port phase images of 200 patients were reconstructed with IRIS and Filtered Back Projection algorithms (FBP). CT number, noise, and Signal-to-Noise Ratio (SNR) were measured for the liver, kidney, spleen, lumbar muscles, vertebral bone, spinal cord, and abdominal aorta. Two experienced radiologists scored the images.

There was no significant difference in CT numbers of the organs under study between FBP and IRIS algorithms (P ˃ 0.05), but the noise difference was significant for all organs (P ˂ 0.05).  The SNR of the aorta and spleen for IRIS was higher than FBP (P ˂ 0.05). The average score of images for FBP and IRIS was 4.12 ± 1.35 and 3.91 ± 0.84, respectively. The Kappa coefficient was reported to be 0.73 with P ˂ 0.01.

The IRIS algorithm does not change the CT number of tissue but reduces the image noise. However, the subjective score of IRIS is less compared to FBP. Optimal use of IRIS in different CT scan protocols requires the collaboration between radiologists and CT technologists.