Guizhi Xu

ORCID: 0000-0002-9637-0051
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • EEG and Brain-Computer Interfaces
  • Neural dynamics and brain function
  • Neuroscience and Neural Engineering
  • Electrical and Bioimpedance Tomography
  • Advanced Memory and Neural Computing
  • Functional Brain Connectivity Studies
  • Transcranial Magnetic Stimulation Studies
  • Flow Measurement and Analysis
  • Acupuncture Treatment Research Studies
  • Blind Source Separation Techniques
  • Brain Tumor Detection and Classification
  • Non-Destructive Testing Techniques
  • Geophysical and Geoelectrical Methods
  • Muscle activation and electromyography studies
  • Neural Networks and Applications
  • Biofield Effects and Biophysics
  • Advanced MRI Techniques and Applications
  • Welding Techniques and Residual Stresses
  • Wireless Power Transfer Systems
  • Energy Harvesting in Wireless Networks
  • Microwave Imaging and Scattering Analysis
  • Advanced Neural Network Applications
  • Magnetic Properties and Applications
  • Gaze Tracking and Assistive Technology
  • Microgrid Control and Optimization

Hebei University of Technology
2016-2025

Liaoning University of Traditional Chinese Medicine
2025

Chinese Academy of Medical Sciences & Peking Union Medical College
2025

Peking Union Medical College Hospital
2025

Shangrao Normal University
2025

Intelligent Health (United Kingdom)
2025

Economic Research Institute
2024

State Grid Corporation of China (China)
2022-2024

North China University of Technology
2023-2024

Xiangtan Electric Manufacturing Group (China)
2024

The classification of brain tumors is a difficult task in the field medical image analysis. Improving algorithms and machine learning technology helps radiologists to easily diagnose tumor without surgical intervention. In recent years, deep techniques have made excellent progress processing However, there are many difficulties classifying using magnetic resonance imaging; first, difficulty structure intertwining tissues it; secondly, due high density nature brain. We propose differential...

10.3390/brainsci11030352 article EN cc-by Brain Sciences 2021-03-10

Despite rapid developments of wearable self-powered sensors, it is still elusive to decouple the simultaneously applied multiple input signals. Herein, we report design and demonstration stretchable thermoelectric porous graphene foam-based materials via facile laser scribing for decoupled strain temperature sensing. The resulting sensor can accurately detect with a resolution 0.5°C gauge factor 1401.5. nanocomposites also explores synergistic effect between components greatly enhance...

10.1038/s41467-024-55790-x article EN cc-by-nc-nd Nature Communications 2025-01-17

To develop an artificial intelligence (AI)-based algorithm which can automatically detect food items from images acquired by egocentric wearable camera for dietary assessment.To study human diet and lifestyle, large sets of were using a device, called eButton, free-living individuals. Three thousand nine hundred containing real-world activities, formed eButton data set 1, manually selected thirty subjects. 2 contained 29 515 research participant in week-long unrestricted recording. They...

10.1017/s1368980018000538 article EN Public Health Nutrition 2018-03-26

The process of diagnosing brain tumors is very complicated for many reasons, including the brain’s synaptic structure, size, and shape. Machine learning techniques are employed to help doctors detect tumor support their decisions. In recent years, deep have made a great achievement in medical image analysis. This paper proposed wavelet autoencoder model named “DWAE model”, divide input data slice as (abnormal) or no (normal). article used high pass filter show heterogeneity MRI images...

10.3390/diagnostics11091589 article EN cc-by Diagnostics 2021-08-31

Atmospheric pressure plasma jets can directly act on the surface of materials and improve properties materials. However, concentration active species generated by is low, treatment area small, which makes it difficult to realize ideal material modification effect within a short processing time. In this paper, method was proposed increase helium jet applying steady-state trapezoidal magnetic field, thus enhancing polyimide (PI) films. Compared case without under 0.25 T field reduced water...

10.1021/acs.langmuir.4c04729 article EN Langmuir 2025-04-06

Brain–computer interface (BCI) is a typical direction of integration human intelligence and robot intelligence. Shared control an essential form combining agents in common task, but still faces lack freedom for the agent. This paper proposes Centroidal Voronoi Tessellation (CVT)-based road segmentation approach brain-controlled navigation by means asynchronous BCI. An electromyogram-based mechanism introduced into BCI system self-paced control. A novel CVT-based method provided to generate...

10.34133/cbsystems.0024 article EN cc-by Cyborg and Bionic Systems 2023-01-01

Tumor detection using medical images plays a key role in practices. One challenge tumor is how to handle the nonlinear distribution of real data. Owing its ability learning data without any prior knowledge, one-class support vector machines (SVMs) have been applied detection. The conventional SVMs, however, assume that each feature sample has same importance degree for classification result, which not necessarily true applications. In addition, parameters SVM and kernel function also affect...

10.1109/tmag.2011.2158520 article EN IEEE Transactions on Magnetics 2011-10-01

Ultrasound scanning has become a highly recommended examination in prenatal diagnosis many countries. The accurate identification of fetal brain ultrasound scans is crucial to head measurement and lesion detection, such as the biparietal diameter detection hydrocephalus. In recent years, deep learning made great progress field image processing. However, there are two difficulties standard planes (FBSPs). First, since tissue not mature, features easy be detected. Second, because expensive...

10.1109/access.2020.2991845 article EN cc-by IEEE Access 2020-01-01

Processing magnetic resonance images are very complex and constantly studied by the researchers to give doctors better ability diagnose patients. In order detect automatically suspicious regions or tumors, we present a new approach inspired threshold segmentation based on morphological operations in this paper. The advantages of our come from complementarities between these two approaches. extract roughly tumor region eventually can affect healthy while method gives clear picture structure...

10.4236/jbise.2016.910b006 article EN Journal of Biomedical Science and Engineering 2016-01-01

Deep learning methods have made some achievements in the automatic skin lesion recognition, but there are still problems such as limited training samples, too complicated network structure, and expensive computational costs. Considering inherent power-efficiency, biological plausibility good image recognition performance of spiking neural networks (SNNs), this paper we make malignant melanoma benign melanocytic nevi lesions classification using convolutional SNNs with unsupervised...

10.1109/access.2020.2998098 article EN cc-by IEEE Access 2020-01-01

Two-dimensional ultrasound scanning (US) has become a highly recommended examination in prenatal diagnosis many countries. Accurate detection of abnormalities and correct fetal brain standard planes is the most necessary precondition for successful measurement. In past few years, support vector machine (SVM) other learning methods have been devoted to automatic recognition 2D ultrasonic images, but performance not satisfactory due wide diversity postures, shortage data, similarities between...

10.1109/access.2019.2950387 article EN cc-by IEEE Access 2019-10-30

Motor imagery (MI) refers to the mental rehearsal of movement in absence overt motor action, which can activate or inhibit cortical excitability. EEG mu/beta oscillations recorded over human cortex have been shown be consistently suppressed during both imagination and performance movements, although specific effect on brain function remains confirmed. In this study, Granger causality (GC) was used construct functional network subjects resting state based order explore effects function....

10.3390/brainsci12020194 article EN cc-by Brain Sciences 2022-01-30

Abstract Doping a small amount of ethanol gas (EtOH) in argon can change the plasma jet from filamentary discharge mode to diffuse mode, and further doping trace oxygen significantly enhance composition content oxygen-containing active particles plasma. Based on this, characteristics Ar + EtOH under different concentrations its effect surface modification polyimide (PI) films were investigated this paper. It was found that O 2 deteriorate with increase concentration, but concentration 0–4000...

10.1088/1361-6595/ad716b article EN Plasma Sources Science and Technology 2024-08-22

WiTricity is a new technology for transmitting energy wirelessly via resonant coupling in the non-radiative near-field. In this paper, design of transmission system implantable devices base on performed, influence structural parameters efficiency investigated, relations between transfer with distance are analyzed, and also frequency values coils different have been figured out.

10.1109/tmag.2011.2174341 article EN IEEE Transactions on Magnetics 2012-01-31

Recently, deep learning methods, in particular, convolutional neural networks (CNNs), have made a massive breakthrough computer vision. And big amount of annotated data is the essential cornerstone to reach this success. However, medical domain, it usually difficult (and sometimes even impossible) get sufficient for some specific tasks. Consequently, work, novel augmentation solution, called semi-supervised attention-guided CycleGAN (SSA-CycleGAN) proposed resolve problem. Specifically,...

10.1109/bibm47256.2019.8982932 article EN 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) 2019-11-01

Non-intrusive load monitoring (NILM) is an important research direction and development goal on the distribution side of smart grid, which can significantly improve timeliness demand response users’ awareness load. Due to rapid development, deep learning becomes effective way optimize NILM. In this paper, we propose a novel identification method based long short term memory (LSTM) learning. Sequence-to-point (seq2point) introduced into LSTM. The innovative combination LSTM seq2point brings...

10.3390/en14030684 article EN cc-by Energies 2021-01-29

Objective: Medical image processing is an exciting research area. In this paper, we proposed new brain tumor detection and classification model using MR images to help the doctors in early of with high performance best accuracy. Materials: The was trained validated five databases, including BRATS2012, BRATS2013, BRATS2014, BRATS2015, ISLES-SISS 2015. Methods: advantage hybrid its novelty that used for first time; our method based on a deep convolution neural network watershed auto-encoder...

10.2174/1573405617666210224113315 article EN Current Medical Imaging Formerly Current Medical Imaging Reviews 2021-02-25
Coming Soon ...