Energy-Threshold Bias Calculator: A Physics-Model-Based Adaptive Correction Scheme for Photon-Counting CT

Area of Science:

• Medical Imaging
• Physics

Background:

• Photon-counting detector computed tomography (PCCT) is advancing rapidly.
• Spectral inconsistency (inter-pixel variations in detected counts) causes artifacts and inaccuracies in PCCT images.

Purpose of the Study:

• To propose a novel physics-model based method to correct spectral inconsistency in PCCT.
• To reduce ring/band artifacts and improve accuracy in reconstructed CT images.

Main Methods:

• Developed a method modeling spectral inconsistency with two terms: a fixed spectral skew term and a variable energy-threshold bias term.
• Introduced an energy-threshold bias calculator (ETB-Cal) for efficient, robust calibration without X-ray fluorescence materials.
• Incorporated correction vectors into a spectral model for pixel-by-pixel correction of PCCT projection data.

Main Results:

• Demonstrated significant reduction in non-uniformity for Gammex multi-energy phantom (29.3 to 5.8 HU).
• Achieved substantial non-uniformity reduction for Kyoto head phantom (27.9 to 3.2 HU).
• Outperformed a polynomial-based method requiring more calibration data.

Conclusions:

• The proposed physics-model based method effectively corrects spectral inconsistency in PCCT.
• ETB-Cal offers computational efficiency, robustness, and convenience for PCCT calibration.
• The method significantly improves image quality by reducing artifacts and non-uniformity.