A5008213343- Non-Invasive Vital Sign Monitoring
- ECG Monitoring and Analysis
- Heart Rate Variability and Autonomic Control
- EEG and Brain-Computer Interfaces
- Hemodynamic Monitoring and Therapy
- Generative Adversarial Networks and Image Synthesis
- Optical Imaging and Spectroscopy Techniques
- Computer Graphics and Visualization Techniques
- Cardiovascular Health and Disease Prevention
- Hand Gesture Recognition Systems
- Arctic and Antarctic ice dynamics
- Hungarian Social, Economic and Educational Studies
- Advanced Computing and Algorithms
- Corrosion Behavior and Inhibition
- Phonocardiography and Auscultation Techniques
- Image Enhancement Techniques
- Traffic Prediction and Management Techniques
- Cryospheric studies and observations
- Fluid Dynamics and Vibration Analysis
- Hydrogen embrittlement and corrosion behaviors in metals
- Structural Integrity and Reliability Analysis
- Video Analysis and Summarization
- Human Pose and Action Recognition
- Russia and Soviet political economy
- Spectroscopy and Chemometric Analyses
Fudan University
2020-2025
Cardiff University
2024
Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou)
2023
Sun Yat-sen University
2023
Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai)
2023
Communication University of Zhejiang
2023
Xi'an Technological University
2023
The proliferation of wearable devices has escalated the standards for photoplethysmography (PPG) signal quality. This study introduces a lightweight model to address imperative need precise, real-time evaluation PPG quality, followed by its deployment and validation utilizing our integrated upper computer hardware system.
Currently, continuous blood pressure (BP) measurements are mostly based on multi-sensor combinations and datasets with limited BP ranges. Besides, most BP-related features derive from the photoplethysmogram (PPG) signal. The mechanism of PPG signal formation is not considered. We aimed to design a noninvasive method for estimation using single sensor, which takes into account.We prepared dataset containing signals 294 patients three public databases constructing model. used in model...
Due to the presence of motion artifacts (MAs), heart rate monitoring using PPG sensors in daily life and physical exercise is challenging, there have been many studies on MA removal algorithms. However, most do not consider quality evaluation signal before removal. In this way, removing directly regardless whether artifact only a waste computing resources, but also easy introduce new noise. paper, detection performed by dividing original into 6s segments calculating amplitude mean difference...
The progression of deep learning and the widespread adoption sensors have facilitated automatic multi-view fusion (MVF) about cardiovascular system (CVS) signals. However, prevalent MVF model architecture often amalgamates CVS signals from same temporal step but different views into a unified representation, disregarding asynchronous nature events inherent heterogeneity across views, leading to catastrophic view confusion. Efficient training strategies specifically tailored for models attain...
Since the twenty-first century, under influence of economic system reform and globalization, Hungary's development has prospered, GDP index increased significantly, however, while country's level year by year, consumer price also shown a increasing trend. Consumer (CPI) is closely related to quality life, employment stability. In this paper, CPI in 2021 taken as base (100), data from 1990 are statistically analyzed, ARIMA model used make time-series forecasts. Based on ADF test,...
Lacking sufficient training samples of different heart rhythms is a common bottleneck to obtain arrhythmias classification models with high accuracy using artificial neural networks. To solve this problem, we propose novel data augmentation method based on short-time Fourier transform (STFT) and generative adversarial network (GAN) evenly distributed in the dataset. Firstly, one-dimensional electrocardiogram (ECG) signals fixed length 6 s are subjected STFT coefficient matrices, then...
With the rise of concept smart cities and healthcare, artificial intelligence helps people pay increasing attention to health themselves. People can wear a variety wearable devices monitor their physiological conditions. The pulse wave is kind signal which widely applied in monitoring system. However, susceptible artifacts, prevents its popularization. In this work, we propose novel beat-to-beat artifact detection algorithm, performs segmentation based on wavelet transform then detects...
A comprehensive investigation, including experimental research, finite-element analysis, and theoretical derivation, was thoroughly carried out on the steel Q370 under strong corrosion to study its fracture mechanism further. Twenty-seven specimens were corroded by 36% industrial hydrochloric acid for 0, 1, 2, 4, 8, 12, 24, 48, 72 h. In particular, a three-dimensional (3D) noncontact laser scanner measured size of pits roughness surface. To sum up, condition that time has leaped, section...
As a physiological signal reflecting the state of muscle activation, surface electromyography (sEMG) plays vital role in assessment neuromuscular health, human–computer interaction, and gait analysis. Inspired by audio analysis outcome that features extracted with Mel Frequency Cepstral Coefficient (MFCC) empower better representation, this paper proposes comparative study gesture recognition method using improving MFCC sEMG. Comparing combining conventional time-domain frequency-domain...
PPG-based blood pressure (BP) estimation has gained great momentum in recent years. However, the commonly used sample pair construction method for training BP model can result weak label problems, which are often obscured by routine supervised learning method. We proposed an end-to-end based on multi-instance and transformer (BP-MIL-Trans). The cropped input PPG segment into several subsegments, with each subsegment defined as instance. Then transformer-based pooling layer could identify...
Blood pressure (BP) estimation based on photoplethysmography (PPG) signals enables continuous and comfortable BP measurement, which is important for the clinical management of hypertension. The purpose this study to propose a novel interpretable feature optimization method improve performance PPG-based model estimation. PPG 152 subjects were selected from public database. Feature detection was performed after FIR band-pass filtering. A total 172 features extracted ten dimensions used...