EOS microdose protocol for the radiological follow-up of adolescent idiopathic scoliosis

Imaging plays a key role in adolescent idiopathic scoliosis (AIS) to determine the prognosis and accordingly define the best therapeutic strategy to follow. Conventional radiographs with ionizing radiation have been associated with 1-2 % increased lifetime risk of developing cancer in children, and physicians, therefore, need a sensitive but harmless way to explore patients at risk, according to the "as low as reasonably achievable" concept. The EOS system (EOS imaging, Paris, France) is available in routine clinical use since 2007, and allows 3D reconstructions of the trunk in standing position with significant radiation reduction. With recent technical advances, further dose reduction can be obtained, but at the cost of image quality that might alter the reliability of 3D reconstructions. The aim of the present study was to analyze the reproducibility of a "microdose" protocol, and evaluate its use in clinical practice.

32 consecutive patients followed for AIS were prospectively included. Biplanar radiographs were obtained with the EOS system according to the new microdose protocol. From the microdose images obtained, three experienced operators performed 3D reconstructions, two times for each subject in a random order (total, 192 reconstructions). The intraoperator repeatability and interoperator reproducibility were evaluated, as recommended by the International Organization for Standardization, for the most clinically relevant 3D radiological parameters.

The identification of the required anatomical landmarks for the "fast spine" reconstruction process was possible in all cases. None of the patients required a second acquisition for 3D analysis. Mean time for reconstruction was 5 ± 2 min. The intraoperator repeatability was better than interoperator reproducibility for all parameters, with values ranging between 3° and 8° for frontal and sagittal spinal parameters, and between 1° and 8° for pelvic measurements. The agreement was very good for all clinical measurements. No correlation was found between the BMI and the reliability of the measurements.

Because children are notably more sensitive to the carcinogenic effects of ionizing radiation, judicious use of imaging methods and a search for newer technologies remain necessary. Results of the current study show that the new microdose acquisition protocol can be used in clinical practice without altering the quality of the images. Relevant clinical measurements can be made manually, but the landmarks are also visible enough to allow accurate 3D reconstructions (ICC >0.91 for all parameters). The resulting radiation exposure was 5.5 times lower than that received with the prior protocol, corresponding now to a 45-fold reduction compared to conventional radiographs, and can, therefore, almost be considered negligible.