New Advances in Electrocardiogram (ECG) Signal Processing

Dear Colleagues,

Over the past decade, there has been significant progress in ECG signal processing, particularly regarding denoising, compression, feature extraction and classification, with these advances resulting in significant improvements to the accuracy, reliability, and clinical relevance of cardiovascular diagnostics.

Traditional signal preprocessing methods, such as band-pass filtering and wavelet transforms, have evolved through adaptive techniques like empirical mode decomposition, variational mode decomposition, and empirical wavelet transform, enabling more precise noise suppression under real-world conditions. Among the most effective approaches are sparse representation methods based on the matching pursuit algorithm, allowing for the decomposition of ECG signals into selective combinations of atoms from redundant dictionaries. The Gabor dictionaries in particular have been leveraged to isolate and suppress noise while preserving critical signal morphology.

At the same time, feature extraction has shifted from time domain (e.g., RR intervals, waveform morphology) and frequency domain spectral features toward robust time frequency representations and data-driven approaches, including deep learning methods using convolutional and recurrent neural networks. Hybrid approaches that combine traditional time frequency transforms with neural networks further improve signal fidelity while balancing computational complexity.

Diagnostic systems based on feature classification have progressed dramatically thanks to machine and deep learning; meanwhile, classic models such as support vector machines, k nearest neighbors, random forests, and ensemble classifiers remain in use. However, CNNs, RNNs/LSTM/BiLSTM, and hybrid CNN LSTM architectures consistently deliver classification accuracies up to 98-99%, especially in arrhythmia detection such as atrial fibrillation and premature ventricular contractions.

The integration of personalization and generalization techniques, such as explainable AI and federated and privacy-preserving learning, personalized adaptation and multi-modal biosignal fusion, are the emerging frontier that promises robust ECG diagnostics capable of adapting to inter-patient variability while preserving data privacy.

Moreover, the integration of real-time signal processing into wearable and mobile platforms has enabled continuous, low-power ECG monitoring, supporting early intervention and telemedicine.

Collectively, these innovations represent a transformative shift toward more intelligent and personalized cardiac care.

Dr. Elzbieta Olejarczyk
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• electrocardiogram
• cardiovascular diseases
• arrhythmia
• signal processing
• denoising
• compression
• feature extraction
• classification
• wavelet transform
• adaptive techniques
• empirical mode decomposition
• variational mode decomposition
• empirical wavelet transform
• matching pursuit
• machine learning
• deep learning
• convolutional and recurrent neural networks
• explainable artificial intelligence
• federated learning
• multi-modal biosignal fusion
• wearable and mobile platforms
• telemedicine
• personalized medicine