Abstract
In this work, a memristive spike-based computing in memory (CIM) system with adaptive neuron (MSPAN) is proposed to realize energy-efficient remote arrhythmia detection with high accuracy in edge devices by software and hardware co-design. A multi-layer deep integrative spiking neural network (DiSNN) is first designed with an accuracy of 93.6% in 4-class ECG classification tasks. Then a memristor-based CIM architecture and the corresponding mapping method are proposed to deploy the DiSNN. By evaluation, the overall system achieves an accuracy of over 92.25% on the MIT-BIH dataset while the area is 3.438 mm2 and the power consumption is 0.178 μJ per heartbeat at a clock frequency of 500 MHz. These results reveal that the proposed MSPAN system is promising for arrhythmia detection in edge devices.
Highlights
Remote healthcare monitoring has received increasing attention for biomedical applications in edge devices
Because such applications are usually deployed in edge devices where computing and memory resources are extremely limited, energy-efficient deep learning (DL) systems are highly required with corresponding software and hardware implementations
To evaluate the performance of the proposed deep integrative spiking neural network (DiSNN), the MITBIH dataset is used in this work (Atzori et al, 2014)
Summary
Remote healthcare monitoring has received increasing attention for biomedical applications in edge devices. Driven by the increasing performance of artificial intelligence (AI), especially deep learning (DL), the healthcare monitoring applications has spread to various aspects including early warning, diagnosis, treatment, and prognosis (Sodhro et al, 2018; Alam et al, 2019; Patan et al, 2020; Pustokhina et al, 2020). Because such applications are usually deployed in edge devices where computing and memory resources are extremely limited, energy-efficient DL systems are highly required with corresponding software and hardware implementations. The main challenge to design such systems is to make the detection of abnormal ECG signals automatic, low-power, and real-time (Yasin et al, 2017; Ai et al, 2018)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
