Electrocardiogram (ECG) acquisition is easily contaminated by interferences, and denoising is the most important task in ECG detection. Variational mode decomposition (VMD) algorithm is widely used in ECG denoising, which can overcome mode aliasing between intrinsic mode function components existed in the traditional empirical mode decomposition (EMD) algorithm, but the mode decomposition number K and penalty factor α in VMD must be optimized to obtain the best signal decomposition accuracy. This paper proposes an improved VMD denoising algorithm that overcomes the shortcomings of slow parameter selection and poor generalization in the traditional VMD algorithm. The algorithm presented firstly adopts EMD algorithm to remove the low-frequency baseline drift noise, and then employs the adaptive particle swarm optimization algorithm to optimize the parameter pair ( K , α) for VMD. To validate the denoising performance of the improved VMD algorithm, the No.103 record from the MIT arrhythmia database is first selected as the pure ECG signal, then both 20dB Gaussian white noises and 0.3 Hz baseline drift are added to simulate the noisy ECG signal. Second, the ECG signals of nine subjects are collected by a customized ECG detection platform based on AD8232 and ADALM1000. The ECG denoising results in simulation and actual experiments show that the improved VMD algorithm achieves the highest signal-to-noise ratio, correlation coefficient and minimum mean square error compared with the traditional EMD and VMD algorithm, which demonstrates the proposed denoising algorithm has stronger denoising ability and better retains morphological characteristics of the original ECG signals.

Voltage sag is one of the most prominent power quality problems in electric power system. Accurate estimation of voltage sag parameters is an important reference for electric utilities and users to evaluate power quality. In this paper, a novel algorithm based on modified S transform with digital prolate spheroidal window (DPSW) is proposed to achieve accurate detection of voltage sag. The DPSW and its time-frequency characteristics are firstly detailed. Then, the construction and performance of modified S transform with DPSW are given. Since DPSW has superior time-frequency distribution, characteristics of voltage sag signal, such as sag depth, duration, and phase jump, can be obtained by analyzing the two-dimensional complex matrix after the modified S transform with DPSW. Simulation and experimental results verify the accuracy and validity of the proposed algorithm.

This study aimed to explore the trend and main influencing factors of road traffic accidents in Guangzhou, China, from 2007 to 2020 and to provide a reference and guidance for government decision-making. A retrospective cohort study was used to describe road traffic accidents in Guangzhou. According to the population types, all people with road traffic accidents were divided into migrant workers and the control population. We divided road users, administrative districts, motorcycle types and injury levels into subgroups to investigate the characteristics of road traffic accidents in Guangzhou. The road traffic accident data were derived from the Guangzhou Public Security Traffic Management Integrated System. The incidence rate of road traffic accidents per 10,000 vehicles in Guangzhou decreased from 36.55 in 2007 to 10.07 in 2012, remained relatively stable at 9.47 in 2017, and finally rose to 11.12 in 2020. The injury rate showed the same trend as the incidence rate, while the mortality rate gradually decreased from 14.21 in 2007 to 5.19 in 2020. Vulnerable road users such as motorized two-to-three-wheeler drivers and migrant workers were casualties in more than 80% of the cases. The proportion of casualties involving mopeds and electric bicycles increased rapidly after 2018. Motor vehicle drivers frequently caused road traffic accidents and were most often uninjured. Road safety in Guangzhou has shown a clear trend of improvement, but casualties are uneven across administrative districts. More attention should be given to motorized two-to-three-wheelers, migrant workers, and road traffic violations by uninjured individuals.

Objective: Modern lifestyles are triggering stress at a disproportionate rate for longer periods of time. Chronic or long-lasting stress can pose a risk to our health. Despite advances in physiological recording methods, mental stress remains challenging to quantify and monitor. Methods: We describe an Internet of Medical Things (IoMT) device with electrocardiogram (ECG) recording features. The recorded ECG signal is processed on-the-fly to calculate, in real time, heart rate (HR), HR variability, energy expenditure, and mental stress. Data are sent to an online platform using a standard Internet of Things (IoT) publish-subscribe messaging transport protocol for continuous monitoring. Results: The system functionality is first validated by performing hardware-in-the-loop measurements connected to a patient simulator. We, then, monitored induced stress by recording ECG in subjects using liquid metal electrodes performing a plank walking task in a virtual reality (VR) environment with high heights exposure. The results demonstrate our IoMT system’s ability to provide accurate ECG metrics using novel liquid metal electrodes by detecting continuously increased stress values in a VR setting and at home. Conclusion: The IoMT measurement device presented provides a novel strategy for monitoring stress in real time. Significance: Our work provides the opportunity for future research on psychological stress and emotion regulation within daily life and the physiological mechanisms through which it influences the health of both children and adults.

Wideband current sources that can work stable over a wide frequency range with large load variations are critical for the performance of electrical impedance spectrum (EIS) measurements. In this paper, we design a wideband mirrored current source (MCS) using two simple differential difference amplifiers (DDA) AD830. PSpice simulation and measurement were performed on load resistances from 100 Ω to 1500 Ω over the frequencies up to 1 MHz. The maximum variation of the output current is less than 0.5% over the whole bandwidth and the output impedance is 4.2 MΩ at 100 Hz and greater than 185 kΩ at 1 MHz. Moreover, less than 1% linearity error (the V _IN versus the I _OUT ) of the output current I _OUT from 1 mA to 20 mA was achieved by sweeping the input voltage signal V _IN from 0.1 V to 2 V at 100 Hz, 1 kHz and 10 kHz respectively. The DDA-based wideband MCS has high accuracy, linearity and simple structure, which may improve the performance of EIS measurements and other relevant applications.

As a new type of functional spectral imaging technique, the multi-frequency electrical impedance tomography (mf-EIT) can reconstruct images according to frequency-dependent conductivity distribution of biological tissues. However, the traditional mf-EIT systems are usually based on stepped-sine (sweeping-frequency) excitation, which have high signal-to-noise ratio (SNR) for recording stationary biological characteristic but are not suitable for capturing dynamic physiological activities because of its relatively long sweeping time. This paper developed a serial 16-electrode mf-EIT system based on multi-frequency simultaneous (MFS) technique, which adopts multisine excitation for real-time spectral imaging at 9 logarithmically distributed frequencies ranging from 10 kHz to 995 kHz with the maximum frame rate of 20 FPS. The hardware and software implementations of the mf-EIT system are described in detail with modular architecture, which have the novel features of a wideband mirrored current source for adjustable MFS excitation with customized spectrum, a highly flexible electrode switching array for arbitrary sensing configuration, integer-period sampling strategy for accurate MFS spectral demodulation, and graphic software for parameter-adjustable real-time mf-EIT spectral imaging. The performance of the developed mf-EIT system is firstly verified on a circle saline (CS) phantom, which shows high SNR and channel consistency. Moreover, frequency-difference EIT (fd-EIT) spectral imaging on a lung-shape raw-potatoes (LSRP) phantom and time-difference EIT (td-EIT) spectral imaging on a two-balloon cooked-mashed-potatoes (TBCMP) phantom are carried out, which show similar imaging performance metrics compared to the traditional mf-EIT systems. As the first serial mf-EIT system that supports real-time spectral imaging, it greatly decreases the system hardware complexity, cost and development difficulty compared to parallel mf-EIT systems, and has significant clinical and industrial application prospects, especially on monitoring dynamic biological or industrial processes.