We also have X and podcasts
Smartphone AI Transforms Intraoperative Diagnosis of Lung Cancer Surgery
A new deep learning model called SuRImage uses smartphone photos of surgical tissue to instantly diagnose lung cancer invasiveness, potentially eliminating 30+ minute waits for traditional analysis while enabling junior surgeons to match senior-level diagnostic performance.
Background
Surgeons treating stage IA lung adenocarcinoma must decide between segmentectomy and lobectomy based on tumor invasiveness. Currently, frozen section analysis provides this information but requires 30+ minutes of processing time during surgery. SuRImage was developed to enable real-time diagnosis using smartphone imagery captured under natural lighting conditions.
Key Findings
In a prospective study of 1,727 patients across three hospitals, SuRImage demonstrated strong diagnostic accuracy:
- AUC of 0.84 (95% CI 0.82-0.86) for invasive adenocarcinoma identification
- AUC of 0.87 for ternary diagnosis (AIS/MIA/IAC classification)
- AUC of 0.85 for multi-grade assessment
- Substantially outperformed frozen section analysis (80.51% vs 22.22% accuracy for grade 1 tumors)
- Junior surgeons assisted by SuRImage achieved 85.83% accuracy—exceeding unassisted senior surgeon performance of 79.10%
Why It Matters
SuRImage enables instant intraoperative guidance, eliminating critical delays while improving diagnostic accuracy. By empowering junior surgeons to achieve senior-level performance, this technology could democratize expertise across institutions and potentially improve surgical outcomes in stage IA lung adenocarcinoma.
Limitations
The study was conducted in China with limited diversity. Clinical implementation requires workflow integration and broader validation. Long-term patient outcomes and cost-effectiveness remain to be established.
Original paper: Deep learning model for pathological invasiveness prediction using smartphone-based surgical resection images in clinical stage IA lung adenocarcinoma (SuRImage): a prospective, multicentric, diagnostic study. — The Lancet. Digital health. 10.1016/j.landig.2025.100965
