Automated health detection of congestive heart failure subject using rank multiresolution wavelet packet attributes and 1-norm linear programming ELM

• Published in: Multimedia tools and applications Vol. 81; no. 14; pp. 19587 - 19608
• Main Authors: Gelmecha, Demissie J., Singh, Ram S., Sinha, Devendra K., Tekilu, Dereje
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.06.2022; Springer Nature B.V
• Subjects: 1205: Emerging Technologies for Information Hiding and Forensics in Multimedia Systems; Age; Artificial neural networks; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Decomposition; Heart failure; Heart rate; Linear programming; Machine learning; Multimedia Information Systems; Neural networks; Ranking; Special Purpose and Application-Based Systems; Support vector machines; Multiresolution Wavelet Packet (MRWP) decomposition; Bhatacharya ranking method; Regularization parameters; Quadrature mirror filter
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-021-11562-z

As far as the mortality of the global population is concerned, it is cardiovascular diseases which cause the highest death rate worldwide, mostly due to the Congestive Heart Failure (CHF). Therefore, an initial detection and diagnosis of CHF becomes essential. This manuscript presents a novel approach to detect health of CHF subject which is based on Multiresolution Wavelet Packet (MRWP) decomposition method, attributes ranking approach, kernel principle component analysis ( KPCA ) and 1 - Norm Linear Programming Extreme Learning Machine ( 1 - NLPELM ) . For this investigation, the heart rate variability (HRV) signal has been decomposed up to 5-level using MRWP decomposition method. The sixty three log root mean square (LRMS) attributes were extracted from the decomposed HRV signal. The top ten attributes are selected by ranking approaches such as Fisher , Wilcoxon, Entropy , Bhattacharya , and receiver operating characteristic ( ROC ) . The ten ranked attributes were then mapped to one new feature by KPCA and fed to 1 - NLPELM . The HRV database of normal subjects (normal sinus rhythm ( NSR ) , age 22–45 years old and elderly (ELY), age 60–82 years old) and CHF subjects (age 32–71 years old) were obtained from PhysioNet ATM. The simulation results demonstrated that Bhatacharya + KPCA with 1 - NLPELM approach achieved an accuracy of 98.44 ± 1.4 % , 99.13 ± 1.85 % for NSR - CHF and ELY - CHF respectively. Out of all ranking methods, Bhatacharya combined with KPCA + 1 - NLPELM provided the highest degree of accuracy for all datasets. In addition, the proposed method has also achieved very good generalization performance and less execution time as compared to 1 - NLPELM , KPCA + PNN ,  KPCA + SVM , probabilistic neural network ( PNN ) and support vector machine ( SVM ) .