Technical Note: FreeCT_wFBP: A robust, efficient, open-source implementation of weighted filtered backprojection for...

Area of Science:

• Medical Imaging
• Computational Imaging
• Image Reconstruction

Background:

• Growing interest in quantitative imaging, radiomics, and computer-aided diagnosis (CAD) using CT necessitates exploring acquisition and reconstruction parameter impacts.
• Existing workflows often lack the scalability for large-scale reconstruction projects, requiring extensive scanner access.

Purpose of the Study:

• To develop and evaluate a freely available, open-source software package for helical fan-beam CT image reconstruction.
• The software implements weighted filtered backprojection (wFBP) to address limitations in large-scale reconstruction.

Main Methods:

• Developed FreeCT_wFBP, a GPU-based, low-dependency reconstruction program using C and Nvidia CUDA C.
• Evaluated reconstruction speed, image quality (CT number accuracy, uniformity, CNR), and mass-attenuation coefficient accuracy using the ACR CT accreditation phantom and a simulated FORBILD thorax phantom.

Main Results:

• Achieved an average reconstruction time of 17.4 ± 1.0 s for a 512x512x32 volume.
• ACR phantom reconstructions met all CT Accreditation Program criteria.
• Reconstructed mass-attenuation coefficients for water agreed with original values within 0.01%.

Conclusions:

• FreeCT_wFBP is a fast, configurable, open-source reconstruction package for third-generation CT.
• The software demonstrates robust performance for both clinical and simulated CT data, supporting quantitative imaging research.