Precision Risk Stratification in Atrial Fibrillation: Evaluating Machine Learning Models for Bleeding Prediction and Clinical Integration
Yazarlar (4)
Arş. Gör. Onur Akgün Türkiye
Arş. Gör. Murat Akdoğan Türkiye
Arş. Gör. Ahmet ARDAHANLI Kafkas Üniversitesi, Türkiye
İsa Ardahanlı
Bilecik Şeyh Edebali Üniversitesi, Türkiye
Bilecik Şeyh Edebali Üniversitesi, Türkiye
| Makale Türü | Özgün Makale (SSCI, AHCI, SCI, SCI-Exp dergilerinde yayınlanan tam makale) | ||
| Dergi Adı | The American Journal of Cardiology (Q2) | ||
| Dergi ISSN | 0002-9149 Dergi Bilgileri (2025) | ||
| Dergi Tarandığı Indeksler | SCI-Expanded | ||
| Makale Dili | Türkçe | Basım Tarihi | 07-2025 |
| Cilt / Sayı / Sayfa | 246 / 0 / – | DOI | 10.1016/j.amjcard.2025.03.015 |
| Makale Linki | https://doi.org/10.1016/j.amjcard.2025.03.015 | ||
| UAK Araştırma Alanları |
Bilgisayar Bilimleri ve Mühendisliği
|
||
| Özet |
| To the Editor, We commend Chaudhary et al. 1 for their insightful study,“Machine Learning Predicts Bleeding Risk in Atrial Fibrillation Patients on Direct Oral Anticoagulant,” published in The American Journal of Cardiology. Their work rigorously compares machine learning algorithms—random forest and XGBoost—to conventional bleeding risk scores (HASBLED, ORBIT, ATRIA) in predicting major bleeding events among atrial fibrillation (AF) patients receiving direct oral anticoagulants (DOACs). The authors report superior discriminative performance of machine learning models (AUC: 0.76) over traditional scores (AUC: 0.57 for HAS-BLED), underscoring machine learning’s potential to refine personalized risk stratification. Notably, their SHAP (SHapley Additive exPlanations) analysis identified novel predictors, such as body mass index, lipid profiles, and insurance type, which may elucidate previously … |
| Anahtar Kelimeler |
BM Sürdürülebilir Kalkınma Amaçları
| Atıf Sayıları | |
| Google Scholar | 7 |