Hybrid iterative reconstruction technique for liver CT scans for image noise reduction and image quality improvement: evaluation of the optimal iterative reconstruction strengths

作者：Song Ji Soo; Lee Jeong Min^*; Sohn Ji Young; Yoon Jeong Hee; Han Joon Koo; Choi Byung Ihn

来源：Radiologia Medica, 2015, 120(3): 259-267.

DOI：10.1007/s11547-014-0441-9

• Hybrid iterative reconstruction
• CT
• Image quality
• Reconstruction algorithm

This study sought to investigate the effect of the hybrid iterative reconstruction (IR) algorithm (iDose, Philips Healthcare) on the improvement of image quality of computed tomography (CT) scans of the liver and determine the appropriate level of IR strength for clinical use.
A total of 75 patients (41 men and 34 women; mean age, 59.5 years) with a primary abdominal malignancy who underwent two-phase liver CT scans for the work-up of their liver metastases, were included in this study. The CT images during the portal phase were reconstructed using either filtered back projection (FBP) or the hybrid IR algorithm with six different levels of IR strengths. The signal-to-noise ratio of the liver (SNRliver) and the contrast-to-noise ratio of the portal vein to muscle (CNRpv to m) were measured. For qualitative analysis, image noise, visibility of small intrahepatic vascular structures, beam-hardening artefact, lesion conspicuity, and overall image quality were graded by two radiologists.
Quantitative analysis demonstrated that image noise was significantly reduced along with the increasing level of iDose and that the values of SNRliver and CNRpv to m were significantly better with iDose than those of FBP images. Qualitative assessment also showed significantly better results with iDose compared with FBP (p < 0.05) and the parameters for subjective image quality were highest with iDose level 4.
The hybrid IR technique is able to reduce image noise and to provide better image quality than FBP, and an intermediate strength of iDose (level 4) provided the highest quality images.

• 出版日期2015-3