AI fares better than doctors at predicting deadly complications after surgery
9 hours ago
- #surgical risk
- #healthcare
- #artificial intelligence
- A new AI model detects signals in routine heart tests (ECGs) that predict post-surgery complications with 85% accuracy.
- The model outperforms current risk scores used by doctors, which are only accurate in about 60% of cases.
- Researchers from Johns Hopkins University developed the model, which combines ECG data with patient medical records for better predictions.
- The AI model could transform how surgical risks are assessed, providing a more accurate tool for patients and surgeons.
- The team plans to further test the model on larger datasets and explore additional information that can be extracted from ECGs using AI.