A5101914251- Advanced MRI Techniques and Applications
- Atomic and Subatomic Physics Research
- Cardiac Imaging and Diagnostics
- Medical Imaging Techniques and Applications
- Higher Education and Teaching Methods
- Text and Document Classification Technologies
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Advanced Neuroimaging Techniques and Applications
- Face and Expression Recognition
- Advanced Measurement and Detection Methods
- Optical Systems and Laser Technology
- MRI in cancer diagnosis
- Advanced Decision-Making Techniques
- Advanced Algorithms and Applications
- Educational Reforms and Innovations
- Complex Network Analysis Techniques
- Customer churn and segmentation
- Advanced Graph Neural Networks
- Innovation and Knowledge Management
- Military Strategy and Technology
- Soil, Finite Element Methods
- Video Coding and Compression Technologies
- Marine and Offshore Engineering Studies
- Remote Sensing and Land Use
Taishan University
2024
Zhoukou Normal University
2012-2023
Zhengzhou University of Aeronautics
2022-2023
China University of Petroleum, East China
2017
Tianjin Open University
2016
Advanced MRI Technologies (United States)
2013-2015
University of California, Berkeley
2013-2015
Advanced Imaging Research (United States)
2010-2012
University of Utah
2010-2012
Yangzhou University
2007-2011
Purpose Simultaneous multi‐slice (SMS) echo planar imaging (EPI) is incorporated into two‐dimensional (2D) arterial spin labeling (ASL) to produce more slices for measuring perfusion in a larger region of the brain than currently possible with EPI. Methods Pulsed ASL (PASL) preparations using FAIR and QUIPSS II techniques were combined SMS‐EPI. Testing was performed four subjects at 3 Tesla. Multiband slice acceleration factors (MB) from MB‐2 MB‐5 40 averages evaluated. Comparisons made...
To determine the feasibility of three-dimensional (3D) hybrid radial (stack-of-stars) MRI with spatiotemporal total variation (TV) constrained reconstruction for dynamic contrast enhanced myocardial perfusion imaging.An ECG-triggered saturation recovery turboFLASH sequence undersampled stack-of-stars sampling TV was developed imaging. Simulations were performed to study dependence approach steady state on flip angle and time this acquisition. Phantom studies used show effect selection...
Abstract Purpose: To develop and test a hybrid radial (stack of stars) acquisition compressed sensing reconstruction for efficient late gadolinium enhancement (LGE) imaging the left atrium. Materials Methods: Two schemes, kx‐ky‐first kz‐first, are tested using signal equation an inversion recovery sequence with simulated data. Undersampled data reconstructions then performed approach three‐dimensional total variation constraint. The framework is on five atrial fibrillation patients after...
Quantitative estimates of myocardial perfusion generally require accurate measurement the arterial input function (AIF). The saturation signal intensity in blood that occurs with most doses contrast agent makes obtaining an AIF challenging. This work seeks to evaluate performance a method uses radial k-space sequence and multiple recovery times (SRT) quantify cardiovascular magnetic resonance (CMR).
Abstract Current myocardial perfusion MRI acquisitions are performed with a saturation recovery sequence, in large part to minimize sensitivity arrhythmia. A new approach is proposed here where the images acquired ungated at steady state without use of pulse. The data continuously and reach after first few images. confluence advances has made this paradigm an steady‐state acquisition possible—very rapid undersampled readouts reconstruction technologies permit enough measurements that...
Understanding the evolutionary patterns of real-world complex systems such as human interactions, biological transport networks, and computer networks is important for our daily lives. Predicting future links among nodes in these dynamic has many practical implications. This research aims to enhance understanding evolution by formulating solving link-prediction problem temporal using graph representation learning an advanced machine approach. Learning useful representations provides greater...
In order to solve the problem that individual coordinates are easily ignored in localization of abnormal behavior marine fish, resulting low recognition accuracy, execution efficiency and high false alarm rate, this paper proposes a method fish based on deep learning network model. Firstly, shadow data is removed, background image subtracted from each frame get gray school. Then, label watershed algorithm used identify different individuals swarm obtained. Combined with experimental size...
Purpose To develop a novel, simultaneous multi‐slice (SMS) reconstruction that extends an inter‐slice leakage constraint to intra‐slice aliasing with virtual slice concept for artifact reduction. Methods Inter‐slice has been used SMS mitigates artifacts from the adjacent slices. In this work, is extended more general framework includes and parallel MRI as special cases by viewing undersampling slices while imposing data fidelity ensure measurement consistency. way, makes it feasible directly...
In view of the complexity multimodal environment and existing shallow network structure that cannot achieve high-precision image text retrieval, a cross-modal retrieval method combining efficient feature extraction interactive learning convolutional autoencoder (CAE) is proposed. First, residual convolution kernel improved by incorporating two-dimensional principal component analysis (2DPCA) to extract features extracting through long short-term memory (LSTM) word vectors efficiently graphic...
To investigate the relationships among highly constrained back projection (HYPR)-LR, reconstruction focal underdetermined system solver (PR-FOCUSS), and k-t FOCUSS by showing how each method relates to a generalized reference image method. That is, series model employs fixed multiplicative corrections-that is extended here consider images more broadly, both in space transform spaces (x-t x-f spaces), that can evolve with iteration.Theoretical between methods were derived. Computer...
Network intrusion detection is a powerful means to identify and analyze the state of Internet things. For reliability requirements things, an analysis method things based on deep network model proposed. First, Inception architecture as backbone network, this constructs multi-scale convolutional neural (M-CNN) model. The long-term short-term memory models are introduced into M-CNN enhance local feature extraction ability At same time, batch normalization global average pooling layers make...
Deck installation is always a major challenge for floating structures, particularly deep draft floaters like the spar which must be installed in relatively water. Derrick barges have been used deck installations until now. The 4000 mt Kikeh Spar was successfully using floatover method November 2006, off coast of Sabah South China Sea. This demonstrates feasibility this concept and opens door more decks on spars future. paper will review technical challenges associated with type installation....
Purpose: To improve rank constrained reconstructions for undersampled multi‐image MRI acquisitions. Methods: Motivated by the recent developments in low‐rank matrix completion theory and its applicability to rapid dynamic MRI, a new reordering‐based reconstruction of data that uses prior image information is proposed. Instead directly minimizing nuclear norm estimated images, reordered values minimized. The reordering based on estimates. method tested brain diffusion imaging contrast...
One of the key MRI methodologies to identify and characterize coronary artery disease is dynamic contrast enhanced myocardial perfusion imaging. Rapid acquisition images can help in improved diagnosis by accurately measuring temporal dynamics injected agent. Another competing requirement complete coverage heart with high spatial resolution better sub-endocardial infarcts reduce dark rim artifacts. Most undersampled reconstruction methods that break this spatio-temporal tradeoff are sensitive...
Introduction Myocardial perfusion MRI is a useful modality to detect myocardial ischemia. Quantitative estimates require an accurate arterial input function (AIF). Recently, method for estimating T1 and thus gadolinium concentration from radial k-space sequence was proposed [1]. The created four sub-images with differing effective saturation recovery times (eSRTs) 96 ray acquisitions estimate T1. No measures of truth were used evaluate the in vivo. In this work, we employ similar technique...
Aiming at the problems about incompleteness and low accuracy in classifying image feature extraction existing object detection methods. First, an improved method based on faster Region-Convolutional Neural Network (RCNN) is proposed, Region of Interest (ROI) align used instead RoI pooling to reduce error process. Second, it adds layer convolution before adding network full-connection layer, thus reducing its parameters, enhance performance classifier avoid over-fitting. Finally, combination...
Many existing image and text sentiment analysis methods only consider the interaction between modalities, while ignoring inconsistency correlation of data, to address this issue, an aspect level multimodal model using transformer multi-layer attention is proposed. Firstly, ResNet50 used extract features, RoBERTa-BiLSTM features. Then, through direct mechanism deep mechanism, multi-level fusion information graphic carried out remove images unrelated given aspect. The emotional representations...
The practical teaching of ideological and political theory courses in colleges universities plays a significant role that is unparalleled by theoretical classroom instruction. current state course instruction, however, exhibits insufficient understanding the essence teaching, resulting lack effectiveness education. It imperative for students to transition from passive reception active participation during teaching. Only through organic integration student-centered instruction evaluation can...