Model-based iterative reconstruction and adaptive statistical iterative reconstruction: dose-reduced CT for detecting...

Area of Science:

• Radiology
• Medical Imaging
• Computational Imaging

Background:

• Iterative reconstruction techniques are crucial for lowering radiation doses in computed tomography (CT).
• Optimizing CT protocols is essential for balancing image quality and patient safety.

Purpose of the Study:

• To evaluate the detectability of pancreatic calcification using ultralow-dose CT (ULDCT) images reconstructed with model-based iterative reconstruction (MBIR) and adaptive statistical iterative reconstruction (ASIR).

Main Methods:

• Prospective study involving 85 patients undergoing unenhanced CT at three radiation dose levels: reference-dose CT (RDCT), low-dose CT (LDCT), and ULDCT.
• Images were reconstructed using filtered-back projection (FBP), ASIR (L-ASIR), and MBIR/ASIR (UL-MBIR, UL-ASIR).
• Two radiologists assessed pancreatic calcification detection using a five-point certainty scale.

Main Results:

• UL-MBIR demonstrated significantly higher sensitivity (0.67-0.89) for detecting pancreatic calcification compared to L-ASIR or UL-ASIR (0.11-0.44).
• The area under the receiver operating characteristic curve (AUC) for UL-MBIR (0.818-0.860) was comparable to L-ASIR (0.696-0.844).
• Specificity was lower with UL-MBIR (0.79-0.92) than with L-ASIR or UL-ASIR (0.96-0.99).

Conclusions:

• Model-based iterative reconstruction (MBIR) enables high-sensitivity detection of pancreatic calcification even at ultralow radiation doses.
• Clinicians should be aware of the potential for slightly decreased specificity when using UL-MBIR for pancreatic calcification detection.