ECG heartbeat classification in compressive domain for wearable devices

The heartbeat classification of ECG signals on wearable devices has attracted extensive attention in recent years. Many existing works have studied them, but they do not consider the energy consumption of wearable device for classification of ECG signals. These methods are thus not suitable for wearable devices. In this paper, we propose a novel ECG heartbeat classification scheme performed in the compressive domain to reduce energy consumption for wearable devices. Specifically, we develop a new QRS detection algorithm that finds the position of the QRS complexes directly on the compressive ECG measurements without signal reconstruction, followed by a deep boltzmann machine (DBM) based classification. Extensive experiments are implemented on MIT-BIH database and our database to verify the proposed scheme. The results show the efficacy of our scheme. More specifically, when CR is 40%, our scheme achieves an accuracy of 90.00% and 81.88% on MIT-BIH database and our database, respectively, compared to an accuracy of 94.38% and 89.38% for the benchmarking method, while ensuring that the energy consumption of wearable devices is reduced. The results also demonstrate that our scheme is effective in significantly improving CPU runtime for wearable devices.