Clinical Image Coregistration Variability on a Dedicated Radiosurgery Unit.

Area of Science:

• Medical physics
• Radiosurgery techniques
• Image-guided therapy

Background:

• Frameless radiosurgery utilizes cone-beam computed tomography (CBCT) for stereotactic definition.
• Magnetic resonance imaging (MRI) is crucial for delineating treatment targets.
• Reproducible coregistration between MRI and CBCT is essential for precise target localization.

Purpose of the Study:

• To assess the reproducibility of image coregistration for frameless radiosurgery.
• To evaluate the accuracy of MRI-to-CBCT coregistration in patient imaging.

Main Methods:

• Three coregistration types (MRI-to-CT, MRI-to-CBCT, CT-to-CBCT) were analyzed in 15 patients.
• Coregistration was performed 5-30 times per type, totaling 465 coregistrations.
• Target Registration Error (TRE) and Compound Registration Error (CRE) were measured to quantify accuracy.

Main Results:

• Median TRE was 0.11 mm (MRI-to-CT), 0.17 mm (MRI-to-CBCT), and 0.12 mm (CT-to-CBCT).
• MRI-to-CBCT showed a statistically significant higher TRE (P < .01).
• Median CRE was 0.44 mm, with submillimeter errors within the central 16 cm of the head.

Conclusions:

• Image coregistration demonstrated high reproducibility, with variations generally within 0.2 mm.
• Compound Registration Error (CRE) is submillimeter in the central 16 cm region, ensuring accuracy for target localization.
• The findings support the reliable use of MRI-to-CBCT coregistration in frameless radiosurgery.