Use of a Hybrid Iterative Reconstruction Technique to Reduce Image Noise and Improve Image Quality in Obese Patients Undergoing Computed Tomographic Pulmonary Angiography

作者：Kligerman Seth^*; Mehta Dhruv; Farnadesh Mahmmoudreza; Jeudy Jean; Olsen Kathryn; White Charles

来源：Journal of Thoracic Imaging, 2013, 28(1): 49-59.

DOI：10.1097/RTI.0b013e31825412b2

• obesity
• pulmonary angiography
• noise
• image quality

Purpose: To determine whether an iterative reconstruction (IR) technique (iDose, Philips Healthcare) can reduce image noise and improve image quality in obese patients undergoing computed tomographic pulmonary angiography (CTPA).
Materials and Methods: The study was Health Insurance Portability and Accountability Act compliant and approved by our institutional review board. A total of 33 obese patients (average body mass index: 42.7) underwent CTPA studies following standard departmental protocols. The data were reconstructed with filtered back projection (FBP) and 3 iDose strengths (iDoseL1, iDoseL3, and iDoseL5) for a total of 132 studies. FBP data were collected from 33 controls (average body mass index: 22) undergoing CTPA. Regions of interest were drawn at 6 identical levels in the pulmonary artery (PA), from the main PA to a subsegmental branch, in both the control group and study groups using each algorithm. Noise and attenuation were measured at all PA levels. Three thoracic radiologists graded each study on a scale of 1 (very poor) to 5 (ideal) by 4 categories: image quality, noise, PA enhancement, and "plastic" appearance. Statistical analysis was performed using an unpaired t test, 1-way analysis of variance, and linear weighted kappa.
Results: Compared with the control group, there was significantly higher noise with FBP, iDoseL1, and iDoseL3 algorithms (P < 0.001) in the study group. There was no significant difference between the noise in the control group and iDoseL5 algorithm in the study group. Analysis within the study group showed a significant and progressive decrease in noise and increase in the contrast-to-noise ratio as the level of IR was increased (P < 0.001). Compared with FBP, readers graded overall image quality as being higher using iDoseL1 (P = 0.0018), iDoseL3 (P < 0.001), and iDoseL5 (P < 0.001). Compared with FBP, there was subjective improvement in image noise and PA enhancement with increasing levels of iDose.
Conclusion: The use of an IR technique leads to qualitative and quantitative improvements in image noise and image quality in obese patients undergoing CTPA.

• 出版日期2013-1