This study obtained approval from the Medical Ethics Committee of the First Affiliated Hospital of the Medical College of Xi'an Jiaotong University, adhering to the ethical standards set forth in the Declaration of Helsinki. Informed consent was obtained from all participants after a detailed explanation of the study's objectives and procedures. The study recruited participants exclusively from Xi'an Jiaotong University and Xi'an University of Electronic Science and Technology, ensuring that none of them had a history of cardiovascular disease.
The current study highlights that vital sign monitoring is an organized and systematic engineering approach[24]. This approach encompasses several components, including underlying hardware collection devices, IoT transmission devices, and data processing modules, enabling its application in various settings such as outpatient monitoring, emergency monitoring, and home monitoring. Figure 5 illustrates the overall research process outlined in this article.
The study presents a portable monitoring system capable of collecting physiological data from users in various scenarios, including hospital monitoring, emergency monitoring, and remote monitoring. This system can collect indicators such as electrocardiogram, blood oxygen saturation, respiratory rate, heart rate, body temperature, blood pressure, and temperature. Data can be transmitted wirelessly to the upper computer platform in the ward through Bluetooth, or to the cloud platform through 5G.
According to the Nyquist sampling theorem, it is necessary to have a sampling frequency that is at least twice the highest frequency of the continuous signal when discretizing it[25]. Therefore, in this study, the electrocardiogram was sampled at 250Hz, the respiratory rate at 250Hz, and the oxygen saturation curve at 125Hz.
Denoising processing
Numerous studies have highlighted that non-invasive collection methods of electrocardiogram, blood oxygen, and respiration may introduce various types of noise, including power frequency interference, baseline drift, and electromyographic interference, during signal acquisition. To address these issues, we employed adaptive filtering to eliminate the noise, separating the frequency domain of the signal and removing noise within specific frequency bands. Additionally, our team utilized a filtering method based on smoothing functions to further denoise the data and accomplish the denoising function.
Data comparison
To ascertain the data's reliability in this experiment, we compared our platform with the SVM-7521 patient monitor manufactured by Japan Optoelectronics Company, which possesses the medical device registration certificate number (20173211298). To minimize experimental errors, two sets of equipment were utilized, positioned on both sides of the patient. We collected the patient's electrocardiogram, blood oxygen, respiratory rate, and blood pressure for a pre- and post-comparison.