Key result
Federated ECG model matches centralized training performance with an AUC of ~0.79.
Why the study?
Sudden heart disease events can be dangerous during driving, requiring methods to detect cardiac risk patterns from single-lead ECG signals while keeping sensitive data inside the vehicle.
Does an edge-cognitive federated learning framework using single-lead ECG improve cardiac risk pattern detection and reduce false alerts compared to centralized training on the PTB-XL dataset?
Does an edge-cognitive federated learning framework using single-lead ECG improve cardiac risk pattern detection and reduce false alerts compared to centralized training on the PTB-XL dataset?
An edge-cognitive federated learning framework using single-lead ECG achieves diagnostic performance comparable to centralized training on the PTB-XL dataset while preserving privacy and reducing false alerts during driving.
Answers are grounded in this paper and cite their sources.
Supports feasibility of privacy-preserving federated ECG analysis; leaves open prospective clinical validation before adoption.
Heart disease events can happen suddenly and may become dangerous during driving. In this paper, we propose an edge-cognitive framework for heart disease monitoring in the Cognitive Internet of Vehicles (CIoV) using ECG signals. The main goal is to detect cardiac risk patterns while keeping sensitive data inside the vehicle. The model uses a single ECG lead, as recorded in a vehicle by steering-wheel electrodes or a chest patch. Our method has three stages. First, we pretrain a 1D ECG encoder with self-supervised masked signal modeling on unlabeled ECG windows. Second, we use federated learning to train a global classifier in multiple vehicles without sharing raw ECG data. Third, we apply lightweight driver personalization to adapt the model to individual ECG patterns. To reduce unnecessary warnings, we add uncertainty-aware alerting based on MC-dropout and a dual-threshold policy. We evaluate the framework on the PTB-XL dataset using a five-class diagnosis task (NORM, MI, STTC, CD, HYP) and a binary task. With lead II, the federated model reaches a Macro-F1 of 0.478 and an AUC of 0.785 in the five-class task, almost the same as centralized training (0.469 and 0.789), and an accuracy of 0.803 in the binary task. Self-supervised pretraining makes federated training faster and gives a small gain when labels are scarce. The dual-threshold policy lowers the false alert rate on normal windows from 0.234 to 0.059, but more risky windows are missed. Overall, the proposed approach supports practical, privacy-preserving heart monitoring for CIoV.
No takes yet. Share an insight, caveat, or question.
Javanmardi et al. (2026) studied Heart disease monitoring (NORM, MI, STTC, CD, HYP). Edge-cognitive framework with self-supervised pretraining, federated personalization, and uncertainty-aware alerting vs. Centralized training was evaluated on Five-class ECG diagnosis performance (Macro-F1 and AUC) and binary classification accuracy (Macro-F1 0.478, AUC 0.785). The federated ECG model achieved a Macro-F1 of 0.478 and AUC of 0.785 on the five-class diagnosis task, comparable to centralized training.
Synapse has enriched 10 related papers. Consider them for comparative context: