Size of the <scp>C7</scp> – <scp>T1</scp> articular processes is overestimated on lateral radiographs when compared to standing cone beam computed tomography

BACKGROUND: Radiography of the caudal cervical vertebral column is commonly performed in horses, but image distortion may occur due to factors such as scapular superimposition, divergent beam, and the distance between anatomy and detector. These variables can falsely suggest enlargement of the articular processes (APs) at the seventh cervical and first thoracic vertebrae (C7-T1). OBJECTIVES: To assess the presence of magnification of C7-T1 APs as determined by AP:vertebral body (VB) height ratio on lateral radiographs compared to cone beam computed tomography (CBCT) as a gold standard, and to explore inter- and intraobserver agreement and correlation of AP height with subjective grades of osteoarthritis (OA) on lateral radiographs and CBCT. STUDY DESIGN: Retrospective method comparison study. METHODS: Horses imaged between 2021 and 2023 with lateral radiographs and standing CBCT of C6-C7 and C7-T1 were included. A single observer measured the complete dataset, while two observers individually measured a randomised subset of 15 cases. OA grading was applied to this subset. Statistical analyses included Shapiro-Wilk test, Student's t test, Wilcoxon Rank Sum test, and ordinal regression. Intra- and interobserver agreement was evaluated using intraclass correlation coefficients (ICC). RESULTS: Ninety-four horses met inclusion criteria. C7-T1 AP:VB ratios were higher on radiographs (1.15 ± 0.13) than CBCT (1.08 ± 0.15, p = 0.002). No differences were observed at C6-C7. Intraobserver reliability was good to excellent for both modalities. Interobserver agreement was moderate on radiographs but excellent on CBCT. Larger AP height at C7-T1 correlated with higher OA grades on CBCT (95% CI 1.470-3.638), but not radiographs (95% CI -0.5467 to 3.771). MAIN LIMITATIONS: Variability in image quality and lack of correlation with postmortem evaluation. CONCLUSIONS: Radiographs overestimate C7-T1 AP size compared to CBCT, potentially leading to false diagnoses. This effect was absent at C6-C7. CBCT measurements correlated better with OA grade, suggesting radiographic interpretation of OA at C7-T1 should be approached cautiously.