Skip to main navigation menu Skip to main content Skip to site footer

AGREEMENT OF KV PLANAR IMAGING AND CBCT FOR PATIENT POSITIONING IN IMAGE-GUIDED RADIOTHERAPY: A MULTI-SITE ASSESSMENT OF SETUP ACCURACY, CLINICAL AGREEMENT, AND BIAS CORRECTION

Abstract

Background: Highly conformal radiotherapy requires precise daily patient positioning to prevent target underdosing or normal tissue toxicity. While CBCT provides comprehensive 3D volumetric data that accounts for soft tissue variations, planar kV imaging is faster and more efficient, making it crucial to evaluate where 2D bony landmarks serve as an acceptable surrogate for 3D localization. Aim: To assess the impact of kV imaging on patient positioning accuracy across diverse treatment sites compared to CBCT, and to evaluate axis-specific setup agreement utilizing 3D displacement magnitudes and clinical tolerance thresholds. Methods:  Using a prospective within-patient design at the University of Nigeria Teaching Hospital, 39 patients were evaluated across breast (N=9), head and neck (N=10), pelvis (N=10), and spine (N=10) cohorts. A total of 392 measurements (77 CBCT and 315 kV) were captured. Same-session paired kV and CBCT scans were taken at fractions 1 and 6 to assess translational couch shifts (vertical, longitudinal, lateral) and 3D vector magnitudes. Analysis was executed via paired t-tests, Bland-Altman limits of agreement, interobserver reliability statistics, and clinical pass rates within a strict ±3{mm} tolerance. Results: Treatment fractional distributions matched perfectly (1:1) between verification imaging and delivered dose fractions. The spine demonstrated perfect pre-calibration agreement (100.00) across all axes with compressed CBCT standard deviations (SD =0.05 {cm} vertical, & 0.07 {cm} longitudinal). Head and neck setups showed excellent agreement (100.00 vertical/lateral), except for a significant longitudinal discrepancy where kV over-shifted by 0.201 {cm} & p=0.002). Breast setups showed moderate agreement but a dropped lateral pass rate (66.67, p=0.011) reflecting tissue compliance. The pelvis exhibited severe disagreement; planar kV heavily understated misalignments compared to CBCT, which revealed large mean shifts (vertical: 0.24 {cm}, lateral: 0.20 {cm} and massive volatile SDs (0.37 - 0.39 {cm), collapsing pre-calibration vertical and longitudinal clinical agreement to a coin-flip 50.00. Conclusion: Positional agreement between modalities is heavily site-dependent. Planar kV imaging achieves clinically acceptable setup accuracy within a ± 3{mm} tolerance for rigid bony anatomy (spine and head/neck). Conversely, planar imaging possesses critical blind spots regarding internal organ motion and soft-tissue deformation in the pelvis, meaning CBCT remains clinically indispensable in this region to prevent geographical target misses.

Keywords

Image-Guided Radiotherapy (IGRT); Patient Positioning; Setup Accuracy; Planar kV Imaging; Cone Beam Computed Tomography (CBCT); Interfraction Variation; Clinical Tolerance Pass Rate.

PDF