Fast CNN-Based Electrocardiogram Signal Quality Assessment Using Fourier Magnitude Spectrum for Resource-Constrained ECG Diagnosis Devices

Automatic assessment of electrocardiogram (ECG) signal quality plays a vital role in reducing false alarms and improving the trustworthiness of unobtrusive health monitoring devices under noisy ECG recordings, which are unavoidable in continuous health monitoring. In this letter, we present an ECG signal quality assessment (ECG-SQA) method based on the Fourier magnitude spectrum as an input to the 1-D convolutional neural network (1-D CNN) with optimal hyperparameters and activation function, which significantly reduces the CNN model size and computational load of resource-constrained devices. On the untrained ECG databases including single-lead and multilead ECG signals having different kinds of P waves, QRS complexes, and T waves (PQRST) morphologies and various kinds of noise sources, the optimal 1-D CNN-based ECG-SQA method had a sensitivity of 99.30%, and specificity of 95.40% for the three convolution layers, three dense layers, and kernel size of
$3\times 1$
. This study demonstrated that optimal parameter selection can reduce computational resources of 52% with the CNN model size of 852 kB and 67697 parameters as compared with other CNN models. Real-time implementation on Raspberry Pi computing shows that the processing time is 124.4
$\pm$
42.5 ms for checking the quality of 5 s ECG signal.