Vasculitis & polymyalgia rheumatica
Exploring the limit of image resolution for human expert classification of vascular ultrasound images in giant cell arteritis and healthy subjects: the GCA-US-AI project
Authors' conclusion
A resolution of 224 × 224 pixels ensures reliable human expert classification and aligns with the input requirements of many common AI-based architectures.
Study context
- Population
- International experts in giant cell arteritis (GCA) ultrasonography classifying US images (temporal artery and branches, axillary artery) as GCA, normal, or not able to classify250 images each
- Participants
- 31 experts
- Comparison
- Ultrasound image classification at resolutions of 32 × 32, 64 × 64, 128 × 128, 224 × 224 and 512 × 512 pixels
- Outcome
- Minimum image resolution for reliable expert classification of US images for presence or absence of GCA
Methodological detailsweb-based exercise, international experts
- Web-based exercise
- International experts
Performance
As reported in the abstract
Classification performance by resolution
| Test or model | Sensitivity | Specificity |
|---|---|---|
| 224 × 224 | 0.767 (95% CI, 0.737-0.796) | 0.862 (95% CI, 0.831-0.888) |
224 × 224
Sensitivity0.767 (95% CI, 0.737-0.796)
Specificity0.862 (95% CI, 0.831-0.888)
Original publication
Exploring the limit of image resolution for human expert classification of vascular ultrasound images in giant cell arteritis and healthy subjects: the GCA-US-AI project. Ann Rheum Dis. Published 13 June 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →