1. The AI-based MRI radiomic model predicted disease-free survival in osteosarcoma with AUCs up to 0.916, outperforming traditional prognostic factors such as tumor stage and volume.
2. Performance remained consistent across 14 tertiary centers and independent test sets, with sensitivity up to 96% and specificity up to 71%.
3. Predictive accuracy was stable regardless of neoadjuvant chemotherapy status, suggesting broad clinical applicability.
4. Radiomic features showed strong correlations with immune biomarkers (CD3, CD8, and CD8/FOXP3 ratio), supporting biologically meaningful imaging signatures.
5. Researchers recommend prospective validation in non-Chinese populations and exploration of hypoxia-related markers such as HIF-1α to refine model generalizability.

Source: BMC Medicine