China 
Africa 
Explore chapters and articles related to this topic
Dispersive Flies Optimisation: Modifications and Application
Published in Adam Slowik, Swarm Intelligence Algorithms, 2020
Mohammad Majid al-Rifaie, Oroojeni M. J. Hooman, Nicolaou Mihalis
The focus of this experiment is detecting abnormalities in the heart function, called arrhythmias that can be encountered in both healthy and unhealthy subjects. The ICU is equipped with monitoring devices that are capable of detecting dangerous arrhythmias, namely asystole, extreme bradycardia, extreme tachycardia, ventricular tachycardia and ventricular flutter/fibrillation.
A review of arrhythmia detection based on electrocardiogram with artificial intelligence
Published in Expert Review of Medical Devices, 2022
Jinlei Liu, Zhiyuan Li, Yanrui Jin, Yunqing Liu, Chengliang Liu, Liqun Zhao, Xiaojun Chen
In recent years, the study of automatic detection of arrhythmia has become a popular research topic in the field of ECG signal analysis. The traditional approach is a rule-based approach, which combines ECG waveform features with expert knowledge to make a comprehensive diagnosis. However, AI methods, especially DL methods, are usually able to make efficient decisions in large and complex datasets [88]. Most of the studies we reviewed were validated using open source databases. Table 3 lists the detailed parameters of these databases. These databases can be applied to heartbeat classification and rhythmic arrhythmia classification (atrial fibrillation, ventricular tachycardia, ventricular flutter, ventricular fibrillation, etc.). It can be seen from Table 3 that the number of records in Challenge 2017 and CPSC 2018 is much more than other databases. Validation of the model on them is closer to the actual clinical application, but their ECG recording time is relatively short. Therefore, a long-term ECG dataset with many patient records in multiple leads is highly anticipated by researchers. The literature on the analysis of arrhythmia with ML and DL techniques has been studied in detail, as shown in Table 4.
Feature extraction of ECG signal
Published in Journal of Medical Engineering & Technology, 2018
Shanti Chandra, Ambalika Sharma, Girish Kumar Singh
Wavelet is a powerful tool to analyse the nonstationary signal due to its multiresolution property. Therefore, in this work, WT is used to detect the R-peak locations. Beat detection performance of the proposed method is examined by Se, +P and Er, using MIT-BIH arrhythmia database. The performance of the proposed rule based algorithms is evaluated using Matlab 2016b. Table 1 and Figures 3–5 demonstrate that the algorithm can detect features of ECG signal accurately. The comparative study of Table 2 shows that the proposed work provides better performance in terms of Se, +P and Er than several existing methods. Examples 1 and 2 clearly demonstrate that this work is very impactful in cardiac disorders’ interpretation. However, there are some limitations of this work such as, it is not suitable for some rhythms, such as ventricular fibrillation and ventricular flutter, since in these rhythms, the frequency components change frequently, and it is applicable only for single lead ECG recording systems.