An ECG classification using DNN classifier with modified pigeon inspired optimizer

Arrhythmia is a form of heart disease in which the regularity of the pulse is changed.ECG data may be analyzed to detect heart-related illnesses or arrhythmias. This paper presents a wrapper feature selection strategy that employs a Pigeon-inspired optimizer(PIO). The modified Pigeon Inspired Optimi...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Multimedia tools and applications Ročník 81; číslo 7; s. 9131 - 9150
Hlavní autoři: Nainwal, Ashish, Kumar, Yatindra, Jha, Bhola
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York Springer US 01.03.2022
Springer Nature B.V
Témata:
1197: Advances in Soft Computing Techniques for Visual Information-based Systems
Accuracy
Algorithms
Arrhythmia
Artificial neural networks
Classifiers
Computer Communication Networks
Computer Science
Data Structures and Information Theory
Feature extraction
Feature selection
Heart diseases
Machine learning
Model accuracy
Multimedia Information Systems
Optimization
Signal classification
Special Purpose and Application-Based Systems
Wavelet transforms
Deep neural network
Wavelet transform
Heartbeat classification
PIO
ISSN:1380-7501, 1573-7721
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Arrhythmia is a form of heart disease in which the regularity of the pulse is changed.ECG data may be analyzed to detect heart-related illnesses or arrhythmias. This paper presents a wrapper feature selection strategy that employs a Pigeon-inspired optimizer(PIO). The modified Pigeon Inspired Optimizer (MPIO) is used to optimize ECG features and the Deep Neural Network (DNN) to classify the ECG signals. In MPIO, the new blood pigeons were introduced to improve the accuracy of the algorithm. Morphological features, wavelet transform coefficients, and R-R interval dynamic features are extracted for classification of ECG signals. After feature extraction, MPIO is used for feature optimization because optimizing the feature plays a key role in developing the model of machine learning, and irrelevant data features degrade model accuracy and enhance model training time. Using optimised features, the DNN classifier is utilised to classify ECG data. The proposed method achieves 99.10% accuracy, 98.90% specificity, and 98.50% sensitivity. Additionally, when compared with other state-of-the-art methodologies, our method of feature selection also exhibited better outcomes.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1380-7501
1573-7721
DOI:10.1007/s11042-021-11594-5