Artificial intelligence (AI) automated radiographic scoring in rheumatoid arthritis: Shedding light on barriers to implementation through comprehensive evaluation
Authors' conclusion
The performance of the AI systems developed for automating radiographic scoring in RA is insufficient to justify use in research or clinical practice.
Study context
- Population
- Patients with RA; training cohort of 157 patients (1470 radiographs); external test datasets from two hospitals589 radiographs, external test sets
- Participants
- 253 patients
- Comparison
- AI automated Sharp van der Heijde scoring vs manual scoring
- Outcome
- Agreement of AI with consensus score
Methodological detailsdeep learning model development, external validation in two test sets
- Deep learning model development
- External validation in two test sets
Performance
As reported in the abstract
| Test or model | ICC | Spearman's correlation coefficient |
|---|---|---|
| AI | 0.03 to 0.27 | 0.16 to 0.55 |
AI
ICC0.03 to 0.27
Spearman's correlation coefficient0.16 to 0.55
Original publication
AI automated radiographic scoring in rheumatoid arthritis: Shedding light on barriers to implementation through comprehensive evaluation. Semin Arthritis Rheum. Published 5 June 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →