The classification of electrocardiogram (ECG) signals is very important for the automatic diagnosis of heart disease. Traditionally, it is divided into two steps, including the step of feature extraction and the step of pattern classification. Owing to recent advances in artificial intelligence, it has been demonstrated that deep neural network, which trained on a huge amount of data, can carry out the task of feature extraction directly from the data and recognize cardiac arrhythmias better than professional cardiologists. This paper proposes an ECG arrhythmia classification method using two-dimensional (2D) deep convolutional neural network (CNN). The time domain signals of ECG, belonging to five heart beat types including normal beat (NOR), left bundle branch block beat (LBB), right bundle branch block beat (RBB), premature ventricular contraction beat (PVC), and atrial premature contraction beat (APC), were first transformed into time-frequency spectrograms by short-time Fourier transform. Subsequently, the spectrograms of the five arrhythmia types were utilized as input to the 2D-CNN such that the ECG arrhythmia types were identified and classified finally. Using ECG recordings from the MIT-BIH arrhythmia database as the training and testing data, the classification results show that the proposed 2D-CNN model can reach an averaged accuracy of 99.00%. On the other hand, in order to achieve optimal classification performances, the model parameter optimization was investigated. It was found when the learning rate is 0.001 and the batch size parameter is 2500, the classifier achieved the highest accuracy and the lowest loss. We also compared the proposed 2D-CNN model with a conventional one-dimensional CNN model. Comparison results show that the 1D-CNN classifier can achieve an averaged accuracy of 90.93%. Therefore, it is validated that the proposed CNN classifier using ECG spectrograms as input can achieve improved classification accuracy without additional manual pre-processing of the ECG signals.

As a typical example of large and complex mechanical systems, rotating machinery is prone to diversified sorts of mechanical faults. Among these faults, one of the prominent causes of malfunction is generated in gear transmission chains. Although they can be collected via vibration signals, the fault signatures are always submerged in overwhelming interfering contents. Therefore, identifying the critical fault’s characteristic signal is far from an easy task. In order to improve the recognition accuracy of a fault’s characteristic signal, a novel intelligent fault diagnosis method is presented. In this method, a dual-tree complex wavelet transform (DTCWT) is employed to acquire the multiscale signal’s features. In addition, a convolutional neural network (CNN) approach is utilized to automatically recognise a fault feature from the multiscale signal features. The experiment results of the recognition for gear faults show the feasibility and effectiveness of the proposed method, especially in the gear’s weak fault features.

The noise cancellation in electrocardiogram (ECG) signal is very influential to distinguish the essential signal features masked by noises. The power line interference (PLI) is the main source of noise in most of bio-electric signals. Digital notch filters can be used to suppress the PLI in ECG signals. However, the problems of transient interferences and the ringing effect occur, especially when the digitization of PLI does not meet the condition of full period sampling. In this paper, to obtain a better cancellation of the PLI, a designing approach, generating adaptive notch filter (ANF) of sharp resolution, is proposed. The proposed method is concise in algorithm and achieves a more comprehensive reduction of the PLI. It only requires on one fast Fourier transform on the input signal. The spectrum correction method, based on the information from the FFT spectrum of the corrupted signal, is utilized to estimate the harmonic parameters of the PLI. The information of a few main lobe spectral bins in the FFT spectrum is merged such that a compensation signal can be synthesized. By subtracting the compensational signal from the original measurement, the PLI within the investigated signal can substantially reduced. A distinguished advantage of the proposed ANF lies in the fact that no parameters are required to be specified, making the algorithm easier to be implemented. The proposed ANF outperforms conventional notch filters because it not only alleviates the undesirable effects but also better preserves the QRS-complex features in the filtered signal.

Atrial fibrillation, the most common sustained arrhythmia, is still a big challenge for researchers in the medical field. Many studies attempt to realize intelligent classification of AF based on deep learning methods. However, many of the studies focused on investigations of relatively simple datasets collected from a relatively small number of subjects. On the other hand, sophisticated preprocessing is usually adopted to analysis the ECG signals. These two factors significantly affect the generalization ability of the trained models for complicated date sets collected from a large number of subjects. In order to address this problem, an improved multi-scale decomposition enhanced residual convolutional neural network is proposed. The proposed method is applied to the large single-lead ECG dataset provided by the PhysioNet/CinC Challenge 2017, and good classification accuracy is suggested by the testing results. In the proposed method, the original ECG record with a large difference in length is resegmented into short samples of 9 seconds. Then, using the derived wavelet frame decomposition, the segmented short samples are decomposed and reconstituted into sub-signal samples of different scales. We trained the fast down-sampling residual convolutional neural networks (FDResNets) with the original short-signal dataset and the reconstructed dataset of each scale. The transfer learning technique is then applied to couple the three FDResNets with good performance into a multi-scale decomposition enhanced residual convolutional neural network (MSResNet). The FDResNet trained by the 0, 9.375 Hz reconstruction dataset achieved the best performance. After six-fold cross-validation, the average test accuracy reached 87.12%, and the average comprehensive F1 score reached 85.29%. The average test accuracy of the multi-scale residual neural network reached 92.1%, and the average overall F1 score reached 89.9%.