Ultra-low-dose chest computed tomography for interstitial lung disease using model-based iterative reconstruction...

Area of Science:

• Radiology and Imaging
• Pulmonary Medicine
• Medical Physics

Background:

• Interstitial lung disease (ILD) diagnosis relies on high-quality CT imaging.
• Ultra-low-dose CT (ULDCT) aims to reduce radiation exposure but may compromise image quality.
• Reconstruction algorithms significantly impact image quality and quantitative analysis in CT.

Purpose of the Study:

• To evaluate the impact of filtered back projection (FBP) and model-based iterative reconstruction (MBIR) on image quality and quantitative analysis for ILD using ULDCT.
• To compare MBIR with lung and conventional settings against FBP at both ultra-low and standard doses.

Main Methods:

• Prospective enrollment of 52 ILD patients undergoing ULDCT and standard-dose CT.
• Reconstruction of ULDCT using FBP (uFBP) and MBIR (uMBIR-Lung, uMBIR-Stnd); standard-dose CT using FBP (sFBP).
• Subjective and objective image quality assessments (noise, SNR, CNR); quantitative analysis of total lung volume (TLV) and %ILD volume (%ILDV) in 32 patients.

Main Results:

• uMBIR-Lung demonstrated superior subjective image quality compared to uMBIR-Stnd, though inferior to sFBP.
• Objective analysis showed significantly reduced noise (SD) and increased SNR/CNR with MBIR reconstructions (uMBIR-Lung, uMBIR-Stnd) compared to uFBP.
• Quantitative measurements showed no significant bias for TLV between ULDCT MBIRs and sFBP. %ILDV was equivalent between uMBIR-Lung and sFBP, but lower for uMBIR-Stnd.

Conclusions:

• MBIR with a lung setting (uMBIR-Lung) provides improved image quality over conventional settings on ULDCT for ILD.
• uMBIR-Lung achieves quantitative measurements (TLV, %ILDV) comparable to standard-dose CT, despite slightly lower image quality.
• MBIR with lung setting is a promising reconstruction technique for optimizing ULDCT in ILD assessment, balancing radiation dose and diagnostic accuracy.