Confirmation of TST with prior-knowledge accuracy via registered CBCT–CT perfusion comparison

Establish, using accurate 3D registration between dynamic CBCT perfusion and dynamic CT perfusion volumes, whether model-based perfusion reconstruction with the time separation technique employing prior-knowledge basis functions yields CBCT perfusion maps that are consistent with CT perfusion maps.

Background

The authors report, for the first time, that model-based reconstruction with TST using prior-knowledge basis functions appears to work accurately for a real dynamic CBCT perfusion scan. However, they emphasize that a rigorous validation requires a direct, voxel-wise comparison against matched CT perfusion maps, which in turn necessitates reliable registration between CBCT and CT volumes.

This explicit need for confirmation frames a validation problem: perform registration-enabled quantitative comparisons to verify the accuracy of TST with prior knowledge on real CBCT perfusion data.