Evaluation of the Bone Reporting and Data System (Bone-RADS) on computed tomography in children with solitary bone lesions

BACKGROUND: The Bone Reporting and Data System (Bone-RADS) has been developed to facilitate and provide a common language for the management of bone tumors in adults. No bone reporting system has been developed to evaluate solitary bone lesions in children. OBJECTIVE: This study aims to evaluate the interrater agreement and diagnostic performance of Bone-RADS in children with solitary bone lesions on computed tomography (CT). MATERIALS AND METHODS: From January 2014 to January 2026, patients <18 years old with bone lesions and a CT available in our picture archiving and communication system (PACS) were identified for eligibility. Five raters evaluated the CT images after a senior radiologist's informative lecture on Bone-RADS. The final diagnosis, accepted as the reference standard, was determined by a multidisciplinary team based on biopsy or at least 2 years of follow-up. Interrater agreement was assessed. A binary Bone-RADS categorization was also used for diagnostic performance evaluation. RESULTS: Seventy-four patients (22 females) with a mean age of 12.5±3.4 (range, 3-17) years old were enrolled in the study. The interrater agreement among all raters was almost perfect for both Bone-RADS and the binary Bone-RADS classification (P<0.001). Except for one non-advanced rater, all individual agreements with the reference standard were substantial in Bone-RADS (P<0.001). Binary Bone-RADS classification showed high accuracy, ranging from 81.1% to 90.5% across all raters. CONCLUSION: Bone-RADS on CT showed high interrater agreement and diagnostic performance in children with solitary bone lesions. Bone-RADS on CT may be used in children to manage solitary bone lesions.