Key result
Novel ECG-based AA-MFF network achieves ~99% accuracy for heart disease diagnosis.
Why the study?
Continuous IoT-based cardiovascular monitoring requires secure data transmission and privacy protection for sensitive patient data.
Does the proposed IoT-enabled AA-MFF deep learning framework improve heart disease diagnosis accuracy from ECG data?
Does the proposed IoT-enabled AA-MFF deep learning framework improve heart disease diagnosis accuracy from ECG data?
Answers are grounded in this paper and cite their sources.
The proposed IoT-enabled ECG framework with dual chaotic encryption and an attention-aware multi-scale feature fusion network achieves high diagnostic accuracy (99.5%) for heart disease.
The Internet of Things (IoT) is a key component of healthcare data to monitor patients, enable preventive care, and aid in diagnosis and prognosis. Healthcare providers can collect information on vital signs, medication compliance, and other health metrics with the help of IoT-enabled devices, which enable prompt healthcare delivery and proactive interventions. Patient data are sensitive by nature, and trust management in terms of data privacy is important. As a result, offering these services is crucial for IoT-based healthcare systems. In order to provide secure data transmission for continuous cardiovascular health monitoring, this paper presents an IoT-aided electrocardiogram (ECG) framework. First, sensors that are affixed to patients are used to gather the raw ECG data. Next, the data are pre-processed using the secondary variational mode decomposition (SVM-D) technique, which efficiently eliminates noise from the data. Additionally, the data are encrypted using the lightweight dual chaotic encryption (L-DCE) technique, which guarantees data integrity during transmission and safeguards patient privacy. Thus, the creation of two distinct chaotic maps, namely the 4-scroll chaotic attractor and the sine-transform chaotic map, demonstrates the system’s dependability. Consequently, doctors can decrypt these encrypted ECG data files for diagnosis purposes and store them on the cloud. Finally, an attention-aware multi-scale feature fusion network (AA-MFF) is designed in order to complete the diagnosis. Consequently, the suggested AA-MFF model is assessed using a variety of existing models, obtaining an impressive 99.5% accuracy, 96.9% PPV, and 98.5% F1-score on the Cleveland heart disease dataset.
No takes yet. Share an insight, caveat, or question.
Golande et al. (2026) studied Heart disease. IoT-aided ECG framework with L-DCE encryption and AA-MFF network vs. Existing models was evaluated on Diagnostic accuracy, PPV, and F1-score. The proposed attention-aware multi-scale feature fusion (AA-MFF) network achieved 99.5% accuracy, 96.9% PPV, and 98.5% F1-score for heart disease diagnosis on the Cleveland dataset.
Synapse has enriched 10 related papers. Consider them for comparative context: