Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
July 31, 2026PLoS ONEOpen Access

Lightweight hybrid CNN-FFT framework detects ECG abnormalities with ~92% accuracy.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Existing deep learning models for ECG interpretation are often computationally intensive and lack predictive reliability assessments, limiting deployment in resource-limited clinical settings.

Does a lightweight hybrid CNN-FFT framework provide accurate and reliable ECG-based cardiac abnormality detection?

Population

Publicly available PTB-XL dataset

Comparison

Lightweight hybrid CNN and FFT framework for binary and five-class multi-label classification

Design

10-fold cross-validation model evaluation study

Key result

A lightweight hybrid CNN-FFT framework for ECG-based cardiac abnormality detection achieved 92.42% accuracy and 97.8% AUC for binary classification, utilizing only 87K parameters.

Authors

AMA. MalikSASelim Aras

Discussion

Loading...

Member takes

Key expert perspectives

Captured external expert commentary on this paper, strongest first. Original sources are linked where available.

R#Reviewer #1Peer Reviewer

“In this work, Malik and Aras propose a hybrid machine learning architecture based on convolutional neural networks and fast Fourier transform features to extract relevant information from ECG signals in a public dataset. They show that this approach achieves strong performance in classifying normal versus abnormal ECG recordings while using a relatively small number of model parameters. The results also suggest that Fourier-based features substantially improve model performance.”

PLOS One
R#Reviewer #2Peer Reviewer

“The authors develop machine learning models for binary and multi-class classification of ECG arrhythmia using the PTB-XL dataset. The models appear to be lightweight and perform well at the task, and the methods are well-described. The authors have addressed all of this reviewer's concerns - I recommend publication.”

PLOS One

Overview

May support efficient ECG models in low-resource settings; hypothesis-generating pending prospective clinical validation.

Key Points

  • The research aims to develop a lightweight framework for reliable ECG-based cardiac abnormality detection that minimizes computational demands while ensuring accurate predictions.
  • Integrated convolutional neural networks and fast Fourier transform for feature extraction
  • Evaluated on PTB-XL dataset using 10-fold cross-validation
  • Assessed model reliability with expected calibration error
  • Achieved 92.42% accuracy and an AUC of 97.8% for binary classification
  • Generated a macro-AUC of 92.46% for five-class multi-label classification
  • Demonstrated low expected calibration error, indicating well-calibrated probability estimates

Structured PICO

Does a lightweight hybrid CNN-FFT framework provide accurate and reliable ECG-based cardiac abnormality detection?

P
Population
Evaluation of a lightweight hybrid CNN-FFT framework for ECG-based cardiac abnormality detection using the publicly available PTB-XL dataset.
I
Intervention
Lightweight hybrid framework integrating Convolutional Neural Networks (CNN) and Fast Fourier Transform (FFT) components
O
Outcome
Binary and five-class multi-label classification performance (accuracy, AUC) and model reliability (Expected Calibration Error)surrogate

A lightweight hybrid CNN-FFT architecture provides highly accurate and well-calibrated ECG abnormality detection while remaining computationally efficient for resource-constrained settings.

Cite This Study

Malik et al. (2026) studied Cardiac abnormalities. Lightweight hybrid framework integrating CNN and FFT was evaluated on Binary classification of cardiac abnormalities. A lightweight hybrid CNN-FFT framework for ECG-based cardiac abnormality detection achieved 92.42% accuracy and 97.8% AUC for binary classification, utilizing only 87K parameters.

synapsesocial.com/papers/6a6c46f4747664a1aa73bf87https://doi.org/10.1371/journal.pone.0354834
View Full Paper
Ask AI
Bookmark
Share