Technical Note: Evaluation of an iterative reconstruction algorithm for optical CT radiation dosimetry.

Area of Science:

• Medical Physics
• Image Reconstruction
• Radiological Imaging

Background:

• Iterative CT reconstruction algorithms offer improved image quality and reduced radiation dose in X-ray CT.
• Noise reduction is critical in optical CT dosimetry for precise analysis and reliable results.

Purpose of the Study:

• Evaluate an iterative CT reconstruction algorithm for gel dosimetry applications.
• Assess the algorithm's performance, particularly for small field dosimetry challenges.

Main Methods:

• Implemented an ordered subsets convex algorithm with total variation minimization (OSC-TV).
• Tested on uniform, finger, and small field gel dosimetry phantoms.
• Compared OSC-TV reconstructions with filtered backprojection using contrast-to-noise-plus-artifact ratio (CANR) and penumbra width.

Main Results:

• OSC-TV demonstrated a 3-5x improvement in CANR compared to filtered backprojection.
• The algorithm preserved steep dose gradient shapes.
• Improper regularization constants can suppress mean values in very small objects.

Conclusions:

• OSC-TV is a suitable algorithm for gel dosimetry.
• Careful selection of the regularization parameter is essential for accurate reconstruction, especially for small objects relative to the field of view.