Adaptive Classification of Impending Human Heart, Kidney and Liver Failures Based on Measurable Blood-Related Parameters Using MIMO HANFA-ART with ACA Algorithms

Akpan, Vincent Andrew and Omotehinwa, Oluwatosin Temidayo and Agbogun, J. B. (2022) Adaptive Classification of Impending Human Heart, Kidney and Liver Failures Based on Measurable Blood-Related Parameters Using MIMO HANFA-ART with ACA Algorithms. Biomedical Sciences, 8 (2). pp. 97-112. ISSN 2575-3924

Text Adaptive classification of impending Human Heart kidney and liver failures Akpan et all 2022.pdf Download (1MB)

Abstract

Heart, kidney and/or liver failure is a life-threatening condition that demands early detection as well as urgent medical attention and diagnosis based on the classification of their respective symptoms. This paper presents the application of multi-input multi-output hybrid adaptive neural-fuzzy algorithm based on adaptive resonant theory (MIMO HANFA-ART) with adaptive clustering algorithm (ACA) for the adaptive classification of the symptoms of impending human heart, kidney and liver failures based on measurable blood-related parameters obtained from hospitals in Akure metropolis of Ondo State, Nigeria. The ACA consist of an adaptive Gustafson and Kessel clustering (AG-KC) algorithm which is initialized by the K-means clustering algorithm. The 7 classes are: (i) Class 1: heart, (ii) Class 2: kidney, (iii) Class 3: liver, (iv) Class 4: kidney and liver, (v) Class 5: heart and liver, (vi) Class 6: heart and kidney, and (vii) Class 7: heart, kidney and liver. A total of 5888 data set with 16 attributes classified into 7 classes for 368 patients collected from 4 hospitals have been used for this investigation. Comparison of the MIMO HANFA-ART with ACA algorithms with neural-fuzzy classifier trained with the modified error back-propagation with momentum (M-EBPM) algorithm shows the efficiency and superior performance of the MIMO HANFA-ART with ACA algorithms for correct classification and prediction of the symptoms of impending heart, kidney and liver failure. MIMO HANFA-ART with ACA algorithms can be adapted and deployed for real-time online prediction and classification of the symptoms of heart, kidney and liver for early detection and medical attention using advanced biomedical electronics instrumentation techniques and Internet-of-Things (IoT) technologies.

Item Type:	Article
Subjects:	Q Science > QM Human anatomy
Divisions:	Faculty of Natural and Applied Sciences
Depositing User:	mrs chioma hannah
Date Deposited:	22 Nov 2022 10:29
Last Modified:	22 Nov 2022 10:29
URI:	http://eprints.gouni.edu.ng/id/eprint/3915

Actions (login required)

View Item