NTU AI Ruler Predicts Atrial Fibrillation Stroke Risk with 90% Accuracy
NQ Score
100/100
AI Summary (NQ-processed)
A research team from National Taiwan University has developed an AI tool, dubbed the 'AI ruler,' that predicts stroke risk from atrial fibrillation with 90% accuracy. This innovation helps clinicians precisely determine who needs anticoagulant medication, reducing unnecessary bleeding risks. The findings were published in 'npj Digital Medicine,' marking a significant step towards personalized precision medicine.
AI analysis data is not yet available.
Frequently Asked Questions
- Q: Who presented the new AI model for predicting stroke risk?
- A: Dr. Lai Chao-lun, Director of Internal Medicine at Hsinchu NTU Hospital, presented it.
- Q: What is the main clinical drawback of using anticoagulants to prevent stroke?
- A: These drugs may increase the risk of bleeding, such as gastrointestinal bleeding or cerebral hemorrhage.
- Q: What dataset was used by the research team to develop the new AI model?
- A: They used 9,511 newly diagnosed atrial fibrillation cases between 2007 and 2016 from NTU Hospital.
- Q: How many cases were used to validate the model at Hsinchu NTU Hospital and Yunlin Branch?
- A: The team used 1,300 cases at Hsinchu NTU Hospital and 1,242 cases at Yunlin Branch for validation.
- Q: What is the accuracy of traditional assessment tools compared to the new AI model?
- A: Traditional assessment tools have an accuracy of about 60 percent, whereas the new AI has 90 percent accuracy.