CORTEXA
← Browse
openalexAlgorithms2026-07-23Cited by 0

Early-Fusion Two-Channel ECG Scalograms and Transfer Learning for Record-Level Classification of Arrhythmia, Congestive Heart Failure and Normal Sinus Rhythm

Esfandiar Khaleghi, Olga Duran, Iosif Mporas, A.T. Augousti

Automatic electrocardiogram (ECG) classification using deep learning is sensitive to data leakage, class imbalance and the way multi-segment decisions are aggregated at record level. This study presents a two-channel time–frequency early-fusion pipeline for classifying ECG recordings into arrhythmia (ARR), congestive heart failure (CHF) and normal sinus rhythm (NSR). The curated PhysioNet-derived ECGData dataset was used, comprising 81 reconstructed two-channel records: 48 ARR, 15 CHF and 18 NSR. Each 512 s ECG record, sampled at 128 Hz, was segmented into 10 s segments with 50% overlap. Continuous wavelet transform scalograms were generated for both ECG channels and horizontally fused into a single image representation for transfer learning with GoogLeNet. Signal-domain augmentation and class-weighted training were applied to improve robustness to small recording variations and class imbalance. To reduce leakage from overlapping segments, all segments from the same record were kept within the same fold during five-fold record-level cross-validation. Record-level predictions were obtained by averaging posterior probabilities across all segments from each record. The proposed early-fusion model achieved 94.83% mean segment-level accuracy and 98.77% aggregate record-level accuracy, correctly classifying 80 of 81 records. Under the same folds and training configuration, single-channel baselines achieved 75.31% and 87.65% aggregate record-level accuracy, indicating that two-channel early fusion improved classification reliability over single-channel ECG modelling. The only record-level error was a CHF recording classified as ARR. Although the results are promising, the CHF class contained only 15 records and the diagnostic classes originated from different source databases; therefore, further validation on larger, clinically verified multi-lead ECG datasets is required.

View free PDFSource page

Related papers

crossrefAlgorithms2026-07-24

A Machine Learning-Based Predictive Model for Maintenance Management of Combustion Engines in the Agricultural Sector

Ivan-Fredy Jaramillo, Walter Orozco-Iguasnia, Rubén Patricio Alcocer Quinteros, Ricardo Rafael Villarroel-Molina, Alejandro Vilcacundo-Chiluisa

Maintenance management of stationary combustion engines in the agricultural sector remains largely manual, increasing the risk of unplanned downtime. This study developed a machine learning-based predictive model to anticipate failures within a 60-day horizon, enabling the transi…

View free PDFSource page
openalexAlgorithms2026-07-23

A Few-Shot Fault Diagnosis Algorithm for Rotating Machinery Based on Time-Frequency Dual-Stream Mamba and Joint Metric Learning

Yan Xu, Boning Wu, Yingnan Gao, Zhenzhen Jin

Rotating machinery plays a critical role in transmission systems, while the scarcity of fault samples and labeled data limits the performance of existing diagnostic methods under few-shot conditions. This paper proposes a few-shot fault diagnosis method for rotating machinery bas…

View free PDFSource page
openalexAlgorithms2026-07-23

On Discovering Discriminative Itemsets Based on Detecting Frequent Itemsets in Succession

Aristotelis Kompothrekas, Basilis Boutsinas

Discovery of Association Rules is one of the most common Data Mining techniques. Contrast data mining is a focused data mining research area for discovering interesting contrast patterns that state the significant differences between datasets, i.e., frequent itemsets in one datas…

View free PDFSource page
crossrefAlgorithms2026-07-08

Comparative Analysis of Machine Learning Algorithms for Malicious Network-Traffic Classification

Byron Wladimir Oviedo-Bayas, Stefany Michelle Perachimba Panezo, Jorge Humberto Guanin-Fajardo, Stalin Daniel Carreño Sandoya

The classification of malicious network-traffic is critical to cybersecurity. However, to the best of our knowledge, no previous studies have performed a comparative analysis of supervised algorithms for classifying malicious traffic, specifically within the network environment o…

View free PDFSource page
crossrefAlgorithms2026-06-19

A Review of Business Analytics, Machine Learning, and Generative Artificial Intelligence Research 2020–2025: Toward Responsible Artificial Intelligence

Arnold Kamis

This review examines the evolving intersections of data analytics, machine learning, and artificial intelligence—terms that have been frequently conflated since 2016 during a period of increased hype and investment. Following recent reviews across areas such as open innovation, s…

View free PDFSource page