Image quality assessment of an iterative reconstruction algorithm applied to abdominal CT imaging

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Filtered back projection (FBP) is a traditional image reconstruction method.
• Iterative reconstruction (IMR) is an advanced technique aiming to improve image quality.

Purpose of the Study:

• To compare image noise and accuracy between IMR and FBP for porcine liver imaging.
• Evaluate different levels of IMR (L1, L2, L3) against FBP.

Main Methods:

• Used porcine liver for 100 acquisitions to simulate human liver imaging.
• Calculated mean, inter-scan standard deviation (SD), and noise power spectrum (NPS) for FBP and IMR images.
• Assessed spatial SD and noise characteristics.

Main Results:

• Spatial SD decreased with increasing IMR levels (FBP: 9.92, IMR-L1: 4.37, IMR-L2: 3.63, IMR-L3: 2.30 HU).
• IMR improved detectability of faint features compared to FBP.
• IMR-L3 showed greater noise suppression than other levels and FBP, particularly at liver edges.

Conclusions:

• IMR maintains structural information and accuracy comparable to FBP.
• Higher levels of IMR, especially level 3, provide superior noise reduction for liver imaging.
• IMR is a promising technique for enhancing diagnostic quality in liver CT scans.