Spondyloarthritis & psoriatic arthritis
Machine-learning derived algorithms for prediction of radiographic progression in early axial spondyloarthritis
Authors' conclusion
Two traditional models predicted radiographic progression as good as the eight ML models tested in this population.
Study context
- Population
- Patients with chronic back pain for <3 years, suggestive of axial spondyloarthritis (axSpA) (French DESIR cohort, 5 years of follow-up)no missing data
- Participants
- 295 patients
- Exposure
- Machine learning models vs traditional models for prediction of radiographic progression
- Outcome
- Radiographic progression at the spine (increase ≥1 point mSASSS/2 years) or sacroiliac joint (worsening ≥1 grade mNY score between 2 visits); model accuracy by 10-fold cv-AUC
Methodological detailsmulticentre, 10-fold cross-validation
- Multicentre
- 10-fold cross-validation
Main finding
As reported in the abstract
10-fold cross-validated AUC for prediction of radiographic progression
| Group | Value | Cross-validated AUC (cv-AUC) |
|---|---|---|
| M2 | 0.79 | 0.79 |
| M3 | 0.78 | 0.78 |
| Generalized Additive Models (GAM) | 0.77 | 0.77 |
| Discrete Bayesian Additive Regression Trees Samplers (DBARTS) | 0.76 | 0.76 |
| Super Learner | 0.74 | 0.74 |
Value
M20.79
M30.78
Generalized Additive Models (GAM)0.77
Discrete Bayesian Additive Regression Trees Samplers (DBARTS)0.76
Super Learner0.74
Cross-validated AUC (cv-AUC)
M20.79
M30.78
Generalized Additive Models (GAM)0.77
Discrete Bayesian Additive Regression Trees Samplers (DBARTS)0.76
Super Learner0.74
Observational data: associations, not effects.
Original publication
Machine-learning derived algorithms for prediction of radiographic progression in early axial spondyloarthritis. Clin Exp Rheumatol. Published 15 April 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →