Model-Based Iterative Reconstruction Allows Lower Dose

September 8, 2014

CT images reconstructed with low dose model-based iterative reconstruction did not compromise image quality.

Physician reviewers were able to identify an equal number of findings from liver computed tomography (CT) images with a 59 percent reduced radiation dose reconstructed with model-based iterative reconstruction (MBIR) as they were from standard-dose adaptive statistical iterative reconstruction (ASIR), according to the results of a recent study.

“We found that clinically reportable findings identified on the images from the stan­dard dose ASIR series were all identi­fied by two separate reviewers on the images from the reduced dose MBIR series,” wrote William P. Shuman, MD, from the department of radiology at the University of Washington School of Medicine, and colleagues in Radiology. “In 94% of the 136 liver findings, subjective conspicuity, subjective spa­tial resolution, and subjective assess­ment of image noise were assigned a score as the same or better with MBIR than with ASIR; none were assigned scores as nonvaluable.”  

According to the background information in the article, ASIR is a partially iterative technique that lowers image noise allowing technicians to obtain better image quality than reconstruction with filtered back projection or to lower radiation dose. In contrast, MBIR is a fully iterative technique that lowers image noise using back and forward projections. The lower noise gained with MBIR can be used to obtain a comparable image quality or to reduce radiation dose.

“However, the different appearance of CT images re­constructed with MBIR may raise con­cerns about depiction of clinically rele­vant findings,” the researchers wrote.

In this study, Shuman and colleagues wanted to compare images taken with reduced radiation dose MBIR and standard dose ABIR. Fifty patients were enrolled in the study and underwent liver CT. Two passes were made in the portal venous phase, one using standard-dose ABIR and one using reduced -dose MBIR.

The images were then reviewed by one physician who scored the ASIR image quality and marked any findings. Two additional reviewers then reviewed the MBIR images and noted if marked ASIR findings were present on the MBIR images, and assigned the images scores for relative conspicuity, spatial resolution, image noise, and image quality. 

The ABIR reviewed, identified and marked 136 liver findings, all of which were identified by the two MBIR reviewers. The MBIR reviewers scored each of the liver findings as equal to or better than the ASIR images for conspicuity in 94.1 percent of MBIR images; for spatial resolution in 96.7 percent of MBIR images, and for image noise 99.3 percent of the MBIR images.

The volume CT dose index for the ASIR images was 15.2 mGy compared with 6.2 mGy for the MBIR images (P<.001). Similarly, the mean size-specific dose estimate for ASIR images was 16.4 mGy compared with 6.7 mGy for MBIR images, a reduction of 59 percent.

“This da­tum suggests that MBIR reconstruction may allow substantial radiation dose reduction in portal venous phase liver CT without compromising depiction of clinically relevant findings and with acceptable or improve image quality,” the researchers concluded.