Low dose CBCT reconstruction via prior contour based total variation (PCTV) regularization: a feasibility study.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Radiotherapy Technology

Background:

• Compressed sensing reconstruction using Total Variation (TV) often smooths edge details.
• Enhancing edge information is crucial for accurate image interpretation in Cone Beam CT (CBCT).

Purpose of the Study:

• To develop a novel Prior Contour based Total Variation (PCTV) method for CBCT reconstruction.
• To improve edge information preservation and reduce artifacts in low-dose CBCT.

Main Methods:

• Edge information was extracted from prior CT scans and mapped to CBCT images.
• A weight map derived from prior contours was used for TV regularization in PCTV.
• The PCTV method was evaluated using phantoms (XCAT, CatPhan) and patient data, comparing it with TV and Edge Preserving TV (EPTV) methods.

Main Results:

• PCTV significantly enhanced edge information for bone, lung vessels, and tumors compared to TV and EPTV.
• In XCAT studies with reduced projections, PCTV achieved lower relative errors (0.3%) and higher edge cross-correlations (0.78).
• PCTV demonstrated robustness to reduced projection numbers and noise, maintaining resolution and preserving low-contrast edges.

Conclusions:

• PCTV effectively preserves edge information while reducing artifacts and noise in low-dose CBCT.
• The PCTV method shows superiority over TV and EPTV in edge enhancement.
• Improved localization accuracy in radiation therapy is a potential benefit of PCTV.