Performance comparison between a deep learning model and spine surgeons in detecting cervical spinal cord compression on radiographs
AI reads standard neck X-rays better than surgeons for detecting spinal cord compression, enabling earlier diagnosis in under-resourced areas.
A deep learning model for binary classification of cervical spinal cord compression on plain radiographs achieved 94.67% accuracy and AUC 0.99, significantly outperforming spine surgeons (69-71% accuracy) with external validation at a second center. The model could enable early detection in resource-limited settings where MRI access is limited, using only standard cervical radiographs.
What the study was
- Study design
- Retrospective multi-institution DL model development and validation
- Population
- Hospitalized patients with cervical spine radiography and MRI
- Sample size
- 720
- Category
- Diagnostics
- Maturity
- Exploratory
- Journal
- Journal of Neurosurgery: Spine
Why it surfaced
DL model substantially outperforms spine surgeons on plain radiographs with external validation. Clinically ready for resource-limited settings where MRI is unavailable. J Neurosurg Spine.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.