Article

Ischemia episode detection in ECG using kernel density estimation, support vector machine and feature selection

School of Information and Communications, Gwangju Institute of Science and Technology 1, , Oryong-dong, Buk-gu, Gwangju, Republic of Korea. .
BioMedical Engineering OnLine (Impact Factor: 1.43). 06/2012; 11(30):30. DOI: 10.1186/1475-925X-11-30
Source: PubMed

ABSTRACT

Myocardial ischemia can be developed into more serious diseases. Early Detection of the ischemic syndrome in electrocardiogram (ECG) more accurately and automatically can prevent it from developing into a catastrophic disease. To this end, we propose a new method, which employs wavelets and simple feature selection.
For training and testing, the European ST-T database is used, which is comprised of 367 ischemic ST episodes in 90 records. We first remove baseline wandering, and detect time positions of QRS complexes by a method based on the discrete wavelet transform. Next, for each heart beat, we extract three features which can be used for differentiating ST episodes from normal: 1) the area between QRS offset and T-peak points, 2) the normalized and signed sum from QRS offset to effective zero voltage point, and 3) the slope from QRS onset to offset point. We average the feature values for successive five beats to reduce effects of outliers. Finally we apply classifiers to those features.
We evaluated the algorithm by kernel density estimation (KDE) and support vector machine (SVM) methods. Sensitivity and specificity for KDE were 0.939 and 0.912, respectively. The KDE classifier detects 349 ischemic ST episodes out of total 367 ST episodes. Sensitivity and specificity of SVM were 0.941 and 0.923, respectively. The SVM classifier detects 355 ischemic ST episodes.
We proposed a new method for detecting ischemia in ECG. It contains signal processing techniques of removing baseline wandering and detecting time positions of QRS complexes by discrete wavelet transform, and feature extraction from morphology of ECG waveforms explicitly. It was shown that the number of selected features were sufficient to discriminate ischemic ST episodes from the normal ones. We also showed how the proposed KDE classifier can automatically select kernel bandwidths, meaning that the algorithm does not require any numerical values of the parameters to be supplied in advance. In the case of the SVM classifier, one has to select a single parameter.

Download full-text

Full-text

Available from: Jinho Park
  • Source
    • "The features used in our method are fewer in number but are more closely related to medical indicators than the previous methods. The method proposed byPark et al. in 2012has also shown high sensitivity of 94.1%[25]. Identical with the method proposed by Park et al., our method is designed with carefully specified features which are highly correlated with the clinical evidences observed in the ECG signals from patients with myocardial ischemia, such as ST elevation or the changes of T waves. "
    [Show abstract] [Hide abstract]
    ABSTRACT: An automatic method is presented for detecting myocardial ischemia, which can be considered as the early symptom of acute coronary events. Myocardial ischemia commonly manifests as ST- and T-wave changes on ECG signals. The methods in this study are proposed to detect abnormal ECG beats using knowledge-based features and classification methods. A novel classification method, sparse representation-based classification (SRC), is involved to improve the performance of the existing algorithms. A comparison was made between two classification methods, SRC and support-vector machine (SVM), using rule-based vectors as input feature space. The two methods are proposed with quantitative evaluation to validate their performances. The results of SRC method encompassed with rule-based features demonstrate higher sensitivity than that of SVM. However, the specificity and precision are a trade-off. Moreover, SRC method is less dependent on the selection of rule-based features and can achieve high performance using fewer features. The overall performances of the two methods proposed in this study are better than the previous methods.
    Full-text · Article · Jan 2016 · Computational and Mathematical Methods in Medicine
  • Source
    • "They have compared their results with MIT/BIH database and obtained a sensitivity of 99.53% and a positive predictivity of 99.73 and a sensitivity of 99.7% and a positive predictivity of 99.68%. Park et al. (2012) proposes a wavelet interpolation filter to remove motion artefacts around the ST-segment of stress ECG. Inoue and Miyazaki (1998) "
    [Show abstract] [Hide abstract]
    ABSTRACT: The aim of this paper is to highlight the biomedical applications of wavelet transform and corresponding research efforts in imaging techniques. A brief introduction of wavelet transform and its properties that are vital for biomedical applications touched by various researchers has been discussed. Application of electrocardiography in terms of QRS complex is briefly spotlighted. Recent survey of wavelet expansion in medical imaging is another facet of this paper, which includes biomedical image denoising, image enhancement and functional neuro-imaging, including positron emission tomography and functional MRI.
    Full-text · Article · Jul 2015 · International Journal of Biomedical Engineering and Technology
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper explains the development of a delineation algorithm for ECG signals and ST segment classification, based on both wavelet transform and support vector machine (SVM), taking advantage of their specific characteristics. The discrete transform with the mother wavelet Daubechies 4 was used to make the pre-processing, signal filtering and QRS complex detection. The selection of the set of coefficients was made according to the energy of each wavelet's decomposition level. The continuous transform was implemented for T and P wave detection.The detection of the onsets and offsets of each of these waves was evaluated using a combination of both types of wavelet transform, allowing the identification of the characteristic components in ECG signal. Samples of different kinds of diseases contained in QT Database were used for the validation. For the QRS complex it was found a sensivility Se=99,8% and a positive predictivity of P+=99,8%; and for P, QRS and T delineation values of sensibility and positive predictivity over 96% were found applied on different morphologies and different leads. For the ST classification with the SVM it was used different kinds of characteristic vectors, it was found a highest sensitivity of 98.8% and a average near to 80%.
    No preview · Conference Paper · Apr 2013
Show more