Adaptive statistical iterative reconstruction technique for radiation dose reduction in chest CT: a pilot study.

Area of Science:

• Radiology
• Medical Imaging
• Computed Tomography

Background:

• Low-dose computed tomography (CT) protocols are crucial for reducing radiation exposure in patients.
• Assessing image quality and lesion detectability at reduced radiation doses remains a significant challenge in thoracic imaging.

Purpose of the Study:

• To compare lesion detection and image quality of chest CT images acquired at varying tube current-time products (40-150 mAs).
• To evaluate image reconstruction techniques, specifically adaptive statistical iterative reconstruction (ASIR) versus filtered back projection (FBP).

Main Methods:

• Chest CT data from 23 patients were acquired at 40, 75, 110, and 150 mAs.
• Image datasets were reconstructed using varying blends of ASIR and FBP (30%, 50%, 70%).
• Radiologists assessed image noise, lesion conspicuity, and diagnostic confidence; objective noise measurements were also performed.

Main Results:

• ASIR demonstrated acceptable noise levels across all tested mAs values (40-150 mAs), unlike FBP which showed unacceptable noise at lower doses (40, 75 mAs).
• Objective noise measurements were significantly lower with ASIR blending compared to FBP (P < .0001).
• No lesions were missed with either reconstruction method, and lesion conspicuity was comparable, though mild texture artifacts were noted with 70% ASIR blending.

Conclusions:

• Chest CT images acquired at 40 mAs using ASIR achieve acceptable image quality.
• ASIR allows for low-dose chest CT acquisition without compromising diagnostic confidence due to substantial artifacts.