Assessment of thoracic disc degeneration using dual-energy CT-based collagen maps
Simon Bernatz,Alexander Tom Hoppe,Leon David Gruenewald,Vitali Koch,Simon S Martin,Lara Engelskirchen,Ivana Radic,Giuseppe Bucolo,Jennifer Gotta,Philipp Reschke,Renate M Hammerstingl,Jan-Erik Scholtz,Tatjana Gruber-Rouh,Katrin Eichler,Thomas J Vogl,Christian Booz,Ibrahim Yel,Scherwin Mahmoudi
DOI: https://doi.org/10.1186/s41747-024-00500-x
2024-08-26
Abstract:Background: We evaluated the role of dual-energy computed tomography (DECT)-based collagen maps in assessing thoracic disc degeneration. Methods: We performed a retrospective analysis of patients who underwent DECT and magnetic resonance imaging (MRI) of the thoracic spine within a 2-week period from July 2019 to October 2022. Thoracic disc degeneration was classified by three blinded radiologists into three Pfirrmann categories: no/mild (grade 1-2), moderate (grade 3-4), and severe (grade 5). The DECT performance was determined using MRI as a reference standard. Interreader reliability was assessed using intraclass correlation coefficient (ICC). Five-point Likert scales were used to assess diagnostic confidence and image quality. Results: In total, 612 intervertebral discs across 51 patients aged 68 ± 16 years (mean ± standard deviation), 28 males and 23 females, were assessed. MRI revealed 135 no/mildly degenerated discs (22.1%), 470 moderately degenerated discs (76.8%), and 7 severely degenerated discs (1.1%). DECT collagen maps achieved an overall accuracy of 1,483/1,838 (80.8%) for thoracic disc degeneration. Overall recall (sensitivity) was 331/405 (81.7%) for detecting no/mild degeneration, 1,134/1,410 (80.4%) for moderate degeneration, and 18/21 (85.7%) for severe degeneration. Interrater agreement was good (ICC = 0.89). Assessment of DECT-based collagen maps demonstrated high diagnostic confidence (median 4; interquartile range 3-4) and good image quality (median 4; interquartile range 4-4). Conclusion: DECT showed an overall 81% accuracy for disc degeneration by visualizing differences in the collagen content of thoracic discs. Relevance statement: Utilizing DECT-based collagen maps to distinguish various stages of thoracic disc degeneration could be clinically relevant for early detection of disc-related conditions. This approach may be particularly beneficial when MRI is contraindicated. Key points: A total of 612 intervertebral discs across 51 patients were retrospectively assessed with DECT, using MRI as a reference standard. DECT-based collagen maps allowed thoracic disc degeneration assessment achieving an overall 81% accuracy with good interrater agreement (ICC = 0.89). DECT-based collagen maps could be a good alternative in the case of contraindications to MRI.