Preoperative MRI-Derived Navigation Reduces Radiation Exposure in Pediatric Spinal Fusion
Abstract
BACKGROUND: Minimizing radiation exposure during pediatric spinal deformity correction is critical due to the cumulative lifetime effects of ionizing radiation. This study uses preoperative MRI-derived synthetic CT data to potentially eliminate preoperative CT scans entirely and reduce the overall radiation load on patients undergoing spinal fusion for adolescent idiopathic scoliosis (AIS). METHODS: A retrospective cohort study was performed on pediatric patients undergoing posterior spinal fusion for AIS with pedicle screw instrumentation at a single tertiary center. Three modalities were compared: fluoroscopy-only guidance, navigation with preoperative CT registration, and navigation with preoperative MRI converted to synthetic CT registration. Demographic, operative, and radiation parameters, including intraoperative fluoroscopy time and total pre- and intraoperative effective dose (mSv), were recorded and analyzed. MRI was converted to CT using machine learning models that translate the signal intensities of an MRI scan into the Hounsfield units (density values) typical of a CT scan. RESULTS: A total of 94 patients were included in 3 groups: fluoro-only (n=39), CT-only (n=28), and MRI-to-CT conversion (n=27). Groups were demographically comparable (age, sex, BMI, preoperative major curve magnitude; all P>0.05). The number of fused levels and estimated blood loss were also comparable (P=0.827, 0.913, respectively). Intraoperative radiation exposure was significantly reduced in both navigation groups compared with the fluoroscopy-only group. Fluoroscopy time decreased from 90.9±28.0 seconds in the fluoroscopy-only cohort to 48.2±21.9 and 47.5±21.7 seconds in the CT-only and MRI-to-CT groups, respectively (P
Fields of Study & Subjects
Spinal Fractures and Fixation Techniques
Radiation Dose and Imaging
Management of metastatic bone disease
Publication details
- Journal / source
- Journal of Pediatric Orthopaedics
- Publisher
- Ovid Technologies (Wolters Kluwer Health)
- Publication date
- 8 April 2026
- Language
- EN