Noise processing and diagnostics of a spintronic-based IOT sensor for magnetocardiography application using SpintronCardioNet

Noise processing and diagnostics of a spintronic-based IOT sensor for magnetocardiography application using SpintronCardioNet

연구 분야: Infrastructure

학회: Multimedia Tools and Applications

Combining the Internet of Things (IoT) with spintronic advances presents a rare chance to address urgent healthcare problems made worse by the COVID-19 pandemic and an aging world population. In this article, an IoT-based solution that goes beyond conventional electrocardiography (ECG) capabilities for remote cardiac monitoring is presented. The solution makes use of sophisticated Magnetic Tunnel Junction (MTJ) sensors. The gadget uses magnetocardiography (MCG) to obtain electromagnetic signals from the heart by utilizing the sensitivity of spintronic technology. This allows the device to overcome the difficulties caused by low-frequency noise in processing MCG signals. The MCG signals obtained from pre-existing ECG recordings are processed and improved using a novel deep-learning algorithm. This model combines a stacked one-dimensional convolutional neural network (1D CNN) with a Deep, Simple Gated Unit. Most importantly, the model's classification layer correctly detects arrhythmia, allowing for prompt treatment of this serious ailment. The method's performance across several publicly available and clinically annotated datasets indicates its resilience and accuracy. These results demonstrate the device's potential to revolutionize remote cardiac care while also advancing medical spintronic. Additionally, they demonstrate the synergistic potential of fusing cutting-edge sensing technology with artificial intelligence (AI) to tackle modern healthcare issues. The presented results, which include an F1 score of 98%, an accuracy of 98%, a precision of 99%, and a recall of 99%, determine the validity and dependability of the suggested technique.