A5101457431- COVID-19 diagnosis using AI
- Metamaterials and Metasurfaces Applications
- Antenna Design and Analysis
- Advanced Antenna and Metasurface Technologies
- Radiomics and Machine Learning in Medical Imaging
- Advanced Chemical Sensor Technologies
- Insect Pheromone Research and Control
- Machine Learning in Healthcare
- Olfactory and Sensory Function Studies
- Advanced Neural Network Applications
- Gold and Silver Nanoparticles Synthesis and Applications
- Plasmonic and Surface Plasmon Research
- COVID-19 Clinical Research Studies
- Infrastructure Maintenance and Monitoring
- AI in cancer detection
- Animal Vocal Communication and Behavior
- Electromagnetic Compatibility and Measurements
- Machine Learning and Algorithms
- MicroRNA in disease regulation
- Energy Harvesting in Wireless Networks
- Speech and Audio Processing
- Music and Audio Processing
- Autonomous Vehicle Technology and Safety
- Advanced Sensor and Energy Harvesting Materials
- Advanced ceramic materials synthesis
Jiangxi University of Traditional Chinese Medicine
2025
Nanjing Institute of Technology
2024
Huazhong University of Science and Technology
2016-2024
Sixth Affiliated Hospital of Sun Yat-sen University
2021
Sun Yat-sen Memorial Hospital
2021
Sun Yat-sen University
2021
Beijing University of Posts and Telecommunications
2020
University of Toronto
2019
Rationale: Some patients with coronavirus disease 2019 (COVID-19) rapidly develop respiratory failure or even die, underscoring the need for early identification of at elevated risk severe illness.This study aims to quantify pneumonia lesions by computed tomography (CT) in days predict progression illness a cohort COVID-19 patients.Methods: This retrospective included confirmed patients.Three quantitative CT features were automatically calculated using artificial intelligence algorithms,...
Background: The coronavirus disease 2019 (COVID-19) has become a worldwide pandemic since mid-December 2019, which greatly challenge public medical systems. With limited resources, it is natural strategy, while adopted, to access the severity of patients then determine treatment priority. However, our work observes fact that condition many mild outpatients quickly worsens in short time, i.e. deteriorate into severe/critical cases. Hence, been crucial early identify those cases and give...
An electronic nose (e‐nose) mimics the mammalian olfactory system in identifying odors and expands human olfaction boundaries by tracing toxins explosives. However, existing feature‐based odor recognition algorithms rely on domain‐specific expertise, which may limit performance due to information loss during feature extraction process. Inspired olfaction, a smart enabled an all‐feature algorithm (AFOA) is proposed, whereby all features gas sensing cycle of semiconductor sensors, including...
Abstract As COVID-19 is highly infectious, many patients can simultaneously flood into hospitals for diagnosis and treatment, which has greatly challenged public medical systems. Treatment priority often determined by the symptom severity based on first assessment. However, clinical observation suggests that some with mild symptoms may quickly deteriorate. Hence, it crucial to identify patient early deterioration optimize treatment strategy. To this end, we develop an early-warning system...
Depressive disorder is a mental characterized by significant and persistent low mood cognitive impairments such as memory loss. Sojae Semen Praeparatum ( SSP ), traditional Chinese medicine capable of dual use in food for medicinal purposes, can significantly reduce the level inflammatory factors, but it not clear whether contains component that produces an ameliorative effect on depressive acting Tryptophan-Kynurenine pathway (TK pathway). This study aimed to explore material basis...
A simple design of an ultrabroadband and high-efficiency reflective linear polarization convertor is presented in the optical frequency region. This formed from planar anisotropic metasurface using a meander wire structure. The numerical simulation demonstrates that conversion ratio over 90% achieved 181 to 506 THz, which equivalent 94.6% relative bandwidth fory -polarized incident lights under normal incidence. Further results show high efficiency can be sustained well when angle increases...
Monocular road segmentation is a fundamental component of autonomous driving. High accuracy has always been focus the community for this task. However, exploration high-speed methods still lacking. This limits prevalence monocular method. In paper, an efficient encoder-decoder network proposed segmenting from single image. Specifically, to achieve inference, we leverage shallow encoder feature extraction and design lightweight decoder resolution recovery. To avoid drop caused by...
Abstract Objective Pure tone audiometry has played a critical role in audiology as the initial diagnostic tool, offering vital insights for subsequent analyses. This study aims to develop robust deep learning framework capable of accurately classifying audiograms across various commonly encountered tasks. Design, Setting, and Participants single‐centre retrospective was conducted accordance with STROBE guidelines. A total 12 518 were collected from 6259 patients aged between 4 96 years, who...
Abstract Metal-dielectric multilayers are versatile optical devices that can be designed to combine the visible transmittance of dielectrics with electronic properties metals for plasmonic and meta-material applications. However, their performances limited by an interfacial absorption often attributed entirely metal surface roughness. Here, we show during deposition AlN/Ag/AlN SiN x /Ag/SiN multilayers, significant diffusion Ag into top dielectric layer form nanoparticles which excite...
A polarization-insensitive and compact metamaterial absorber (MMA) was proposed in S band, which could exhibit a perfect absorption for electromagnetic (EM) energy harvesting application. The unit cell design consists of an electric-LC (ELC) resonator structure loaded with resistors backed by complementary patch. extracted effective EM parameters simulated surface current distributions the designed MMA indicate that mainly originates from impedance match local magnetic resonance. incident is...
Smart Electronic Nose A smart electronic nose is enabled by an all-feature olfactory algorithm (AFOA) utilizing the response, equilibrium, and recovery processes in a gas sensing cycle. The deep learning-enabled codesign of sensor array recognition can reduce heavy demand for huge amount highly specialized sensors provide interpretable insights into odor dynamics iterative way. More infomration be found article number 2200074 Xiang Bai, Huan Liu, co-workers.
Recently, over-parameterized neural networks have been extensively analyzed in the literature. However, previous studies cannot satisfactorily explain why fully trained are successful practice. In this paper, we present a new theoretical framework for analyzing which call feature repopulation. Our analysis can empirical success of two level that by standard learning algorithms. key result is limit infinite number hidden neurons, two-level via (noisy) gradient descent learns well-defined...
Abstract The ongoing epidemic of SARS‐CoV‐2 is taking a substantial financial and health toll on people worldwide. Assessing the level duration neutralizing antibody (Nab) would provide key information for government to make sound healthcare policies. Assessed at 3‐, 6‐, 12‐, 18‐month postdischarge, we described temporal change IgG levels in 450 individuals with moderate critical COVID‐19 infection. Moreover, data imputation framework combined novel deep learning model was implemented...
We propose a reflective linear polarization converter with broadband and high-efficiency using ladder-shaped resonator (LSR) structure metamaterial (MM) based on multi-reflection interference theory. Three resonances generated by LSRs contribute to the expansion of working frequency range while theory is employed explain characteristic conversion. Both simulation calculation results show that average conversion ratio more than 90% from 7.07 14.3 GHz, which are consistent experimental...
This paper proposes a novel design method for 77 GHz automotive radar antenna. In this method, dual-port network is used to perform full-wave simulation on each antenna element composed of 1/4 wavelength transmission line, 1/2 radiation patch, and line in the array. can greatly reduces error caused by discontinuity patch line. By changing length width element, electric current feeding into conforms Chebyshev distribution. Therefore, half-power beam (HPBW) side lobe level (SLL) required...
Abstract BackgroundRecurrent implantation failure (RIF) is an obstacle in the process of assisted reproductive technology (ART). At present, there was limited research on its pathogenesis, diagnosis and treatment methods.ResultsIn this study, a series analytical tools were used to analyze differences miRNAs, mRNAs lncRNAs endometrium patients RIF group control group. same time, multiple databases are predict target genes non-coding RNAs. Then competing endogenous RNA (ceRNA) network built...
Metal-dielectric interfaces of various geometries are fundamental photonic material platforms for surface plasmons. Surface plasmon polaritons and localized plasmons two modes that excited on planar curved metal-dielectric interfaces, respectively. Herein, we demonstrate in-situ formation silver nanoparticles by diffusion at a interface between sputter-deposited nitride dielectrics. In one step, synthesize nanostructure exhibits both resonances the interface. We develop an accurate optical...
Cong Fang#, Hua-Yao Li#, Long Li, Hu-Yin Su, Jiang Tang, Xiang Bai* and Huan Liu*C. FangSchool of Electronic Information Communications, Huazhong University Science Technology, Wuhan 430074, ChinaH.-Y. L. H.-Y. J. H. Liu School Optical Information, Optics Valley Laboratory, ChinaX. BaiSchool Artificial Intelligence Automation, ChinaE-mail: xbai@hust.edu.cn; huan@hust.edu.cn#These authors contributed equally: C. Fang, LiKeywords : electronic nose, all-feature extraction, deep learning, odor...