Multi-task deep learning model estimates rib fracture age with 7.94-day error on CT

Deep learning model on chest CT estimated rib fracture age with a mean absolute error of 7.94 days vs 10.65 days for manual reads, in external testing of 1,848 fractures.

• The multi-task model also classified healing stage (accuracy 71.18%) and fracture type (accuracy 94.71%).
• Trained on 1,848 fractures from multicenter chest CTs and demonstrated generalizability in external testing.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.