Ultra-low dose abdominal MDCT: using a knowledge-based Iterative Model Reconstruction technique for substantial dose...

Area of Science:

• Radiology
• Medical Imaging
• Image Reconstruction

Background:

• Reduced dose (RD) CT examinations are crucial for minimizing radiation exposure.
• Iterative Model Reconstruction (IMR) is a novel technique for CT image processing.
• Assessing IMR's efficacy in RD abdominal CT is essential for clinical practice.

Purpose of the Study:

• To evaluate lesion detection and image quality of IMR in RD abdominal CT.
• To compare IMR with filtered back projection (FBP) in RD abdominal CT.
• To assess IMR's performance across different patient body mass indices (BMI).

Main Methods:

• Prospective study involving 82 RD abdominal CT exams and 41 patients.
• RD CT data reconstructed using FBP and IMR, compared to standard dose (SD) FBP.
• Radiologists assessed subjective image quality and artifacts on a five-point scale.
• Objective image noise and noise spectral density (NSD) were analyzed.
• A low-contrast phantom study was conducted.

Main Results:

• All lesions present in SD FBP images were detected in RD IMR images.
• RD FBP images showed unacceptable subjective image quality.
• RD IMR provided acceptable image quality for organ margins, soft-tissue structures, and lymphadenopathy, especially in patients with BMI ≤ 25 kg/m².
• Objective image noise was significantly lower with RD IMR compared to RD FBP.
• NSD analysis confirmed reduced noise in the frequency domain with IMR.

Conclusions:

• IMR significantly improved objective and subjective image quality in RD abdominal CT.
• IMR facilitated lesion detection comparable to standard dose CT.
• IMR is a promising technique for optimizing image quality in reduced radiation settings.