Delara analyzes equine radiographs and detects the most clinically relevant structural patterns in fetlock imaging — with high-confidence classification and fully automated objective measurements.
Built on thousands of annotated radiographs, Delara v1.0 focuses on robust detection of early degenerative changes, structural bone remodeling, and normal radiographic appearances.
Delara reliably classifies both abnormal findings and radiographs without significant pathology (WNL) — essential for triage and longitudinal monitoring.
(Performance indicator based on current validation results)
| Supported label | Current indicator |
|---|---|
| Within Normal Limits | +++ strong |
| Kissing spine | +++ strong |
| Spur and lipping | ++ moderate |
| Increased synovial invaginations | ++ moderate |
| Sidebone | ++ moderate |
| Bone remodeling and change | + limited |
| Lucency and subchondral change | + limited |
| Fracture and fragment | + limited |
Delara includes a fully automated segmentation and landmark-detection system providing consistent, reproducible geometric measurements.
Delara automatically compares current radiographs with prior studies in your account. It detects whether findings have:
This enables structured longitudinal monitoring for rehabilitation, post-treatment follow-up, training management, and preventive screening.
Veterinarian feedback directly improves the model. For every detected label, you can:
Your input feeds the training pipeline, accelerating the next generation of equine radiographic AI.