| Abstrakt: |
• The main contributions of this study are listed as follows. • Innovatively use three-lead beat in series and the memory ability of GRU network, to obtain more inter-lead and intra-lead time correlation information. The comprehensive information is more scientific and significant for IMI detection. • The proposed model retains both deep and shallow features of ECG, eliminates the problem of information loss in the process of convolution. These features include more detailed information of IMI. • Through the feature connection, the advantages of the two networks complement each other, help the model obtain more timing information and more comprehensive features. These multi-dimensional features enrich the description of ECG signals, and help the network learn more essential characteristics of IMI, so as to enhance the generalization ability of the model. • We show that without changing model architecture or parameters, the performance of this method remains stable on different datasets, thus having better scalability and applicability. Early and accurate detection of inferior myocardial infarction (IMI) is important for reducing the risk of mortality from a heart attack. Although previous work has demonstrated IMI detection, the differences among patients have been ignored. Most models display excellent performance in the intra-patient scheme, but the inter-patient test results are not ideal. The present paper proposes a model based on densely connected convolutional and gated recursive unit (GRU) networks to enhance the generalization ability of the model. Firstly, the data of multi-lead beat in series is used with GRU, to obtain more inter-lead and intra-lead time correlation information. This correlation information is scientific and significant for IMI detection. Secondly, the proposed model retains both deep and shallow features of ECG through DenseNet, which include more detailed information of IMI. Finally, through the feature connection, the multi-dimensional features enrich the description of ECG signals, and help the network learn more essential characteristics of IMI, so as to enhance the generalization ability of the model. The proposed method was verified by the PTB diagnostic database of the German National Metrology Institute. The generalization ability of the model was tested by intra-patient and inter-patient schemes. After 5-fold cross-validation, the average accuracy, sensitivity and specificity were 99.95%, 99.94% and 99.96% in the intra-patient scheme respectively. Furthermore, these parameters were 88.68%, 90.33% and 87.04% in the inter-patient scheme. The experimental results show that the model displays good generalization ability, which has important clinical significance. [ABSTRACT FROM AUTHOR] |
Full Text from ScienceDirect