3 Key Takeaways
Plain English Summary- 01Core Finding
Images were independently assessed for NBF by 2 readers blinded to other imaging and clinical information.
- 02Study Evidence
Tested in a observational cohort study at the real-world cohort stage.
- 03Practical Meaning
Observational cohort findings reveal real-world links, though lifestyle or unmeasured factors cannot be fully ruled out without an RCT.
What This Means for You
Using ldCT as a reference standard, sCT demonstrated substantially higher sensitivity than T1wMRI and radiography. Intra- and inter-reader reliability and inter-modality agreement between ldCT and sCT were comparable, suggesting that discrepancies in lesion detection primarily reflect reader variability rather than inter-modality differences.
The Discovery
This study aims to compare magnetic resonance imaging (MRI)-based synthetic computed tomography (sCT), T1-weighted magnetic resonance imaging (T1wMRI), low-dose CT (ldCT), and radiography for detecting spinal new bone formation (NBF) in participants with axial spondyloarthritis (axSpA), psoriatic arthritis (PsA), rheumatoid arthritis (RA) and healthy controls (HCs). All participants underwent whole-spine ldCT and MRI in addition to cervical/lumbar spine radiography (not HCs). sCT was reconstructed using BoneMRI. Images were independently assessed for NBF by 2 readers blinded to other imaging and clinical information. One reader repeated ldCT and sCT reads after 6 months. Sensitivity and specificity were calculated using ldCT as the reference standard. Inter-reader and intra-reader reliability were assessed using intraclass correlation coefficient (ICC) and Cohen's kappa. Sixty-nine participants (36 males, mean age 51.4 years) were included: 30 axSpA, 19 PsA, 10 RA, and 10 HCs. The mean number of NBFs per participant was 25.8 on ldCT, 24.0 on sCT, 9.3 on T1wMRI, and 6.9 on radiography. Sensitivity for NBF detection was highest for sCT (70%), followed by radiography (50% cervical/lumbar and a clinically meaningful level whole-spine) and T1wMRI (28%); specificity was ≥93% across modalities. Inter-reader reliability was high for sCT and ldCT (excellent ICCs, substantial kappas), whereas radiography and T1wMRI showed good ICCs with substantial/moderate kappas. Intra-reader reliability and inter-modality agreement for sCT and ldCT were excellent/substantial (ICCs/kappas).
How It Works, Simply
This immunology research focuses on inflammation. Researchers evaluated it using an observational cohort study at the real-world cohort stage of medical research.
3 Key Takeaways
Clinical Takeaways- 01Core Finding
Images were independently assessed for NBF by 2 readers blinded to other imaging and clinical information.
- 02Study Evidence
Tested in a observational cohort study at the real-world cohort stage.
- 03Practical Meaning
Observational cohort findings reveal real-world links, though lifestyle or unmeasured factors cannot be fully ruled out without an RCT.
Key Numbers
Biological Mechanism & Study Design
Study Design & Cohort
Observational cohort study
Observational cohort study · Real-World Cohort
Biological Pathway & Modality
Inflammation
Intervention: Performance of MRI-based synthetic CT for detecting spinal new bone formation in
Experimental Findings & Endpoints
This study aims to compare magnetic resonance imaging (MRI)-based synthetic computed tomography (sCT), T1-weighted magnetic resonance imaging (T1wMRI), low-dose CT (ldCT), and radiography for detecting spinal new bone formation (NBF) in participants with axial spondyloarthritis (axSpA), psoriatic arthritis (PsA), rheumatoid arthritis (RA) and healthy controls (HCs).
All participants underwent whole-spine ldCT and MRI in addition to cervical/lumbar spine radiography (not HCs). sCT was reconstructed using BoneMRI. Images were independently assessed for NBF by 2 readers blinded to other imaging and clinical information.
One reader repeated ldCT and sCT reads after 6 months. Sensitivity and specificity were calculated using ldCT as the reference standard.
Inter-reader and intra-reader reliability were assessed using intraclass correlation coefficient (ICC) and Cohen's kappa. Sixty-nine participants (36 males, mean age 51.4 years) were included: 30 axSpA, 19 PsA, 10 RA, and 10 HCs.
The mean number of NBFs per participant was 25.8 on ldCT, 24.0 on sCT, 9.3 on T1wMRI, and 6.9 on radiography. Sensitivity for NBF detection was highest for sCT (70%), followed by radiography (50% cervical/lumbar and a clinically meaningful level whole-spine) and T1wMRI (28%); specificity was ≥93% across modalities.
Inter-reader reliability was high for sCT and ldCT (excellent ICCs, substantial kappas), whereas radiography and T1wMRI showed good ICCs with substantial/moderate kappas. Intra-reader reliability and inter-modality agreement for sCT and ldCT were excellent/substantial (ICCs/kappas).
Clinical Translation & Practice Context
Using ldCT as a reference standard, sCT demonstrated substantially higher sensitivity than T1wMRI and radiography.
Intra- and inter-reader reliability and inter-modality agreement between ldCT and sCT were comparable, suggesting that discrepancies in lesion detection primarily reflect reader variability rather than inter-modality differences.
The Bottom Line · Clinical Pearl
“Using ldCT as a reference standard, sCT demonstrated substantially higher sensitivity than T1wMRI and radiography.”
Published in Annals of the Rheumatic Diseases · Current Stage: Real-World Cohort
