Key result
Deep neural network predicts ACT status during AF catheter ablation with ~81% accuracy.
Why the study?
The study investigates activated clotting time monitoring and individualized heparin dosing during atrial fibrillation catheter ablation.
Can machine learning models accurately predict activated clotting time (ACT) response and recommend individualized heparin doses during atrial fibrillation catheter ablation?
Observational (n=1,144)
No
Can machine learning models accurately predict activated clotting time (ACT) response and recommend individualized heparin doses during atrial fibrillation catheter ablation?
Machine learning models, particularly deep neural networks, demonstrate feasibility in predicting intraoperative ACT responses and estimating individualized heparin doses during AF ablation, though prospective clinical validation is required.
Answers are grounded in this paper and cite their sources.
May aid initial heparin dosing in AF ablation without altering practice yet; leaves open prospective validation of ML-guided anticoagulation.
This study investigates activated clotting time monitoring and individualized heparin dosing during atrial fibrillation catheter ablation. Data from 1175 patients treated at the First Affiliated Hospital with Nanjing Medical University between January 2020 and December 2022 were collected; after eligibility screening and preprocessing, 1144 cases were analyzed. Random Forest, XGBoost, CatBoost, DNN, LightGBM, and TabNet were evaluated using five-fold cross-validation, with each case serving once in a held-out validation fold. In the six-model comparison, DNN achieved a mean accuracy of 0.81, a Kappa coefficient of 0.68, macro-recall of 0.75, and macro-F1 of 0.72. In a separate five-fold DNN class-balancing analysis, the unbalanced mean accuracy was 0.786 and standard-category recall was 0.485; SMOTE-DNN achieved a mean accuracy of 0.812, Kappa of 0.683, macro-recall of 0.758, macro-F1 of 0.752, and standard-category recall of 0.654. For initial-dose agreement, the mean error was 728 IU, R-squared was 0.68, and 70% of predictions were within 1000 IU of the administered initial dose. In a separate matched ablation involving a 63-case post-outcome-defined subgroup, residual-based sample exclusion reduced dose MAE from 4008 to 3627 IU and increased the 1000-IU hit rate from 17.5 to 23.8%; however, R-squared remained negative, indicating limited absolute accuracy for subsequent-dose recommendation.
No takes yet. Share an insight, caveat, or question.
Sun et al. (2026) conducted an observational in Atrial fibrillation (n=1,144). Deep Neural Network (DNN) model vs. Other machine learning models was evaluated on Accuracy of activated clotting time (ACT) status prediction. A Deep Neural Network model predicted activated clotting time status during atrial fibrillation catheter ablation with a mean accuracy of 0.81 and a Kappa coefficient of 0.68.
Synapse has enriched 10 related papers. Consider them for comparative context: