146. A novel vertebral numbering system for identifying congenital anomalies of the spine using full spine radiographs with the EOS imaging system

Abstract

<h3>BACKGROUND CONTEXT</h3> The numbering of vertebral transitional anatomy can be challenging and controversial. Disagreement exists on how to best report anomalous vertebra and which imaging modality (CT, MRI, X-ray, etc) is most consistent. For these reasons, there is often a discrepancy in the numbering of transitional levels between radiologists and surgeons. Moreover, correlating levels identified intraoperatively by fluoroscopy with inconsistently numbered vertebrae on preoperative MRI or CT is arduous. Errors in numbering can have devastating implications for patients. A simple and accurate vertebral numbering system that can be applied to routine imaging is needed. <h3>PURPOSE</h3> The purpose of this study was to describe a novel and more simplified vertebral numbering system that can be used amongst surgeons and radiologists. <h3>STUDY DESIGN/SETTING</h3> A retrospective imaging study analyzing full-length spine X-rays was performed with the approval of a Colorado Multiple Institutional Review Board (COMIRB protocol:17-1920). <h3>PATIENT SAMPLE</h3> A total of 5,422 images were obtained that included all adult patients who presented to the Spine Center at the University of Colorado and obtained full-spine X-rays using the EOS imaging system between January 31st of 2018 and March 31st of 2020. Images from the same patient at multiple time points were excluded and 3,250 unique patient images were included in our study. After the initial period of data collection, 103 images were excluded from the final analysis due to poor image quality or missing image files. <h3>OUTCOME MEASURES</h3> Total presacral mobile segments were identified and counted using 24 as the number found in conventional spinal anatomy. The total number of ribs were counted for each patient including hypoplastic ribs, and the level of the first-ribbed vertebrae and last-ribbed vertebrae were recorded. <h3>METHODS</h3> Image analysis was conducted by five experienced researchers. The results were reviewed and verified by the senior authors. Any disagreements were reviewed and resolved by two board-certified musculoskeletal radiologists. Final results were evaluated and a novel vertebral numbering system with the lowest probability of error was agreed upon. <h3>RESULTS</h3> In our study cohort, 91.1% (2,868/3,147) demonstrated the conventional 24 presacral mobile segments, whereas 3.3% (104) had 23 presacral mobile segments, 5.5% (174) had 25, and 0.03% (1) had 26. Transitional anatomy, as defined by a number of presacral segments other than 24, was noted in 8.8% (279) of the total number of patients. Within this same group, the majority (94.6%; 2,976) had 12 ribs (including hypoplastic ribs), 4.5% (143) had 11 ribs, and 0.88% (28) had 13 ribs. Just 3.5% (112) had incomplete or hypoplastic ribs. Furthermore, 1.8% (56) of patients had both transitional anatomy and an abnormal number of ribs, with the pattern of transitional anatomy and 11 ribs being the most common combination identified. Considering the first nonribbed vertebrae as the first lumbar vertebrae yielded the lowest probability of a numbering error. <h3>CONCLUSIONS</h3> We are proposing a novel vertebral numbering system that yields the lowest probability of error in reporting anomalous anatomy which can be applied to routine cervical, thoracic, and lumbar radiographs. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.