Entropy-based reliable non-invasive detection of coronary microvascular dysfunction using machine learning algorithm

Aug 4, 2026

Upload by Himbo

Coronary microvascular dysfunction (CMD) is recognized as a key contributor to myocardial ischemia, yet dependable early detection remains limited by the lack of robust non-invasive approaches. This study develops an electrocardiogram (ECG)-based machine learning framework grounded in entropy metrics extracted from ST-T segments. Vectorcardiography (VCG) is computed from 10-second ECG recordings, and sample entropy (SampEn), approximate entropy (ApEn), and complexity index (CI) from multiscale entropy are evaluated. Sequential backward feature selection and comparative modeling identify an optimal classifier, with results showing strong performance and supporting ECG/VCG entropy features for reliable non-invasive CMD detection.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages26

Reading Time1 h 6 min

AccessFree to download

Introduction

Methods

Data and signal processing

Entropy feature extraction

Feature selection and model training

Results

Discussion

Conclusion

Entropy-based reliable non-invasive detection of coronary microvascular dysfunction using machine learning algorithm Zhao, X., Gong, Y., Xu, L., Xia, L., Zhang, J., Zheng, D., Yao, Z., Zhang, X., Wei, H., Jiang, J., Liu, H. & Mao, J. Published PDF deposited in Coventry University’s Repository Original citation: Zhao, X, Gong, Y, Xu, L, Xia, L, Zhang, J, Zheng, D, Yao, Z, Zhang, X, Wei, H, Jiang, J, Liu, H & Mao, J 2023, 'Entropy-based reliable non-invasive detection of coronary microvascular dysfunction using machine learning algorithm', Mathematical Biosciences and Engineering, vol. 20, no. 7, pp. 13061-13085. [https://dx.doi.org/10.3934/mbe.2023582](https://dx.doi.org/10.3934/mbe.2023582) [DOI 10.3934/mbe.2023582](DOI 10.3934/mbe.2023582)[ ](DOI 10.3934/mbe.2023582)[ISSN 1547-1063](ISSN 1547-1063)[ ](ISSN 1547-1063)[ESSN 1551-0018](ESSN 1551-0018) Publisher: AIMS Press This is an open access article distributed under the terms of the Creative Commons Attribution License ( [http://creativecommons.org/licenses/by/4.0](http://creativecommons.org/licenses/by/4.0) ) MBE, 20(7): 13061−13085 . DOI: 10.3934/mbe.2023582 Received: 29 January 2023 Revised: 20 May 2023 Accepted: 22 May 2023 Published: 05 June 2023 [http://www.aimspress.com/journal/MBE](http://www.aimspress.com/journal/MBE) Research article Entropy-based reliable non-invasive detection of coronary microvascular dysfunction using machine learning algorithm Xiaoye Zhao1,2,3,†, Yinlan Gong4,†, Lihua Xu5, Ling Xia6,7, Jucheng Zhang8, Dingchang Zheng9, Zongbi Yao10, Xinjie Zhang10, Haicheng Wei2, Jun Jiang11, Haipeng Liu9,* and Jiandong Mao1,2,3,* 1 School of Instrument Science and Opto-electronic Engineering, Hefei University of Technology, Hefei 230009, Anhui, China 2 School of Electrical and Information Engineering, North Minzu University, Yinchuan 750001, Ningxia, China 3 Key Laboratory of Atmospheric Environment Remote Sensing of Ningxia, Yinchuan 750001, Ningxia, China 4 Institute of Wenzhou, Zhejiang University, Wenzhou 325000, Zhejiang, China 5 Hangzhou Linghua Biotech Ltd, Hangzhou 310009, Zhejiang, China 6 Key Laboratory for Biomedical Engineering of Ministry of Education, Hangzhou 310009, Zhejiang, China 7 Institute of Biomedical Engineering, Zhejiang University, Hangzhou 310009, Zhejiang, China 8 Department of Clinical Engineering, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou 310009, Zhejiang, China 9 Research Centre for Intelligent Healthcare, Coventry University, Coventry, CV1 5FB, United Kingdom 10 Department of Cardiology, Ningxia Hui Autonomous Region People’s Hospital, Yinchuan 750021, Ningxia, China 11 Department of Cardiology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou 310009, Zhejiang, China * [Correspondence:](Correspondence: Email: haipeng.liu@coventry.ac.uk)[ Email: haipeng.liu@coventry.ac.uk](Correspondence: Email: haipeng.liu@coventry.ac.uk), [mao_jiandong@163.com](mao_jiandong@163.com) ; Tel: +44784642- 4479, +8613895003915; Fax: +8609512066815. † These two authors contributed equally. Abstract: Purpose: Coronary microvascular dysfunction (CMD) is emerging as an important cause of myocardial ischemia, but there is a lack of a non-invasive method for reliable early detection ofCMD. Aim: To develop an electrocardiogram (ECG)-based machine learning algorithm for CMD detection that will lay the groundwork for patient-specific non-invasive early detection of CMD. Methods: Vectorcardiography (VCG) was calculated from each 10-second ECG of CMD patients and healthy controls. Sample entropy (SampEn), approximate entropy (ApEn), and complexity index (CI) derived from multiscale entropy were extracted from ST-T segments of each lead in ECGs and VCGs. The most effective entropy subset was determined using the sequential backward selection algorithm under the intra-patient and inter-patient schemes, separately. Then, the c

FAQs

Why is non-invasive early detection of coronary microvascular dysfunction (CMD) important?

CMD is an emerging cause of myocardial ischemia, but existing approaches lack reliable non-invasive methods for early detection, motivating the study’s ECG-based strategy.

What entropy features are extracted from ECG/VCG signals in the proposed method?

The study extracts sample entropy (SampEn), approximate entropy (ApEn), and a complexity index (CI) derived from multiscale entropy, computed from ST-T segments of each ECG lead and corresponding VCG.

Which machine learning model and feature set performs best under the intra-patient scheme?

An ApEn-based SVM model is reported as the optimal choice under the intra-patient scheme, achieving testing evaluation metrics over 0.8.

Aug 4, 2026

Upload by Himbo