Multilevel hybrid accurate handcrafted model for myocardial infarction classification using ECG signals

Abstract

Myocardial infarction (MI) is detected using electrocardiography (ECG) signals. Machine learning (ML) models have been used for automated MI detection on ECG signals. Deep learning models generally yield high classification performance but are computationally intensive. We have developed a novel multilevel hybrid feature extraction-based classification model with low time complexity for MI classification. The study dataset comprising 12-lead ECGs belonging to one healthy and 10 MI classes were downloaded from a public ECG signal databank. The model architecture comprised multilevel hybrid feature extraction, iterative feature selection, classification, and iterative majority voting (IMV). In the hybrid handcrafted feature (HHF) generation phase, both textural and statistical feature extraction functions were used to extract features from ECG beats but only at a low level. A new pooling-based multilevel decomposition model was presented to enable them to create features at a high level. This model used average and maximum pooling to create decomposed signals. Using these pooling functions, an unbalanced tree was obtained. Therefore, this model was named multilevel unbalanced pooling tree transformation (MUPTT). On the feature extraction side, two extractors (functions) were used to generate both statistical and textural features. To generate statistical features, 20 commonly used moments were used. A new, improved symmetric binary pattern function was proposed to generate textural features. Both feature extractors were applied to the original MI signal and the decomposed signals generated by the MUPTT. The most valuable features from among the extracted feature vectors were selected using iterative neighborhood component analysis (INCA). In the classification phase, a one-dimensional nearest neighbor classifier with ten-fold cross-validation was used to obtain lead-wise results. The computed lead-wise results derived from all 12 leads of the same beat were input to the IMV algorithm to generate ten voted results. The most representative was chosen using a greedy technique to calculate the overall classification performance of the model. The HHF-MUPTT-based ECG beat classification model attained excellent performance, with the best lead-wise accuracy of 99.85% observed in Lead III and 99.94% classification accuracy using the IMV algorithm. The results confirmed the high MI classification ability of the presented computationally lightweight HHF-MUPTT-based model.

Keywords: ECG signal processing; Local binary pattern; MI classification; Statistical feature extraction.

Fig. 1

Lead III waveforms on example ECG signals of different classes in the dataset

Fig. 2

Schema of the proposed HHF-MUPTT-based ECG beats classification model

Fig. 3

Schema of the MUPTT. By using all A and M subbands, a pooling band structure has been created to extract features

Fig. 4

Schema of the feature extraction process using 1D-ISBP. Here, v defines the values of the overlapping blocks to extract features, and the center defines the center value of the overlapping block

Fig. 5

Confusion matrix of Lead III classification results. The enumerated classes are: healthy (0), anterior (1), anterior lateral (2), anterior septal (3), inferior (4), inferior lateral (5), inferior posterior (6), inferior posterior lateral (7), lateral (8), posterior (9) and posterior lateral (10). In addition, the numbers of ECG beats in every myocardial infarct class are given in Table 2

Fig. 6

Lead-wise classification accuracies using 75:25 split ratio

Fig. 7

Accuracy rates of the calculated ten voted results

Fig. 8

Confusion matrix of the best voted result

Fig. 9

The informative feature rates for all couples by calculating the t-test

Fig. 10

Overview of the presented alternative model.