The prime purpose of this project is to develop a portable cardiac abnormality monitoring device which can drastically improvise the quality of the monitoring and the overall safety of the device. While a generic, low cost, wearable battery powered device for such applications may not yield sufficient performance, such devices combined with the capabilities of Artificial Neural Network algorithms can however, prove to be as competent as high end flexible and wearable monitoring devices fabricated using advanced manufacturing technologies. This paper evaluates the feasibility of the Levenberg-Marquardt ANN algorithm for use in any generic low power wearable devices implemented either as a pure real-time embedded system or as an IoT device capable of uploading the monitored readings to the cloud.
翻译:本项目的主要目的是开发一种便携式的心脏异常监测设备,可以大幅提高监测质量和整个设备的安全性。虽然对于此类应用而言,通用的、低成本、可穿戴的电池供电设备可能无法产生足够的性能,但是结合人工神经网络算法的能力,这样的设备可以证明与使用高端的柔性和可穿戴监测设备,这些设备使用先进制造技术制造的监测设备是同样有竞争力的。本文评估了Levenberg-Marquardt人工神经网络算法在任何通用低功耗可穿戴设备中的可行性,这些设备可以作为纯实时嵌入式系统或可上传监测读数到云端的物联网设备实现。