A Deep Learning Based Anti-aliasing Self Super-Resolution Algorithm for MRI

2018 ◽  
pp. 100-108 ◽  
Author(s):  
Can Zhao ◽  
Aaron Carass ◽  
Blake E. Dewey ◽  
Jonghye Woo ◽  
Jiwon Oh ◽  
...  
Keyword(s):  
Deep Learning ◽  
Super Resolution ◽  
Resolution Algorithm

scholarly journals Magnetic Resonance Imaging Features under Deep Learning Algorithms in Evaluated Cerebral Protection of Craniotomy Evacuation of Hematoma under Propofol Anesthesia

2021 ◽  
Vol 2021 ◽  
pp. 1-11
Author(s):  
Deqian Xin ◽  
Zhongzhe An ◽  
Juan Ding ◽  
Zhi Li ◽  
Leyan Qiao
Keyword(s):  
Magnetic Resonance Imaging ◽  
Deep Learning ◽  
Magnetic Resonance ◽  
Diffusion Tensor ◽  
Super Resolution ◽  
Control Group ◽  
Imaging Features ◽  
Resonance Imaging ◽  
Resolution Algorithm ◽  
Experimental Group

This study aimed to explore the value of magnetic resonance imaging (MRI) features based on deep learning super-resolution algorithms in evaluating the value of propofol anesthesia for brain protection of patients undergoing craniotomy evacuation of the hematoma. An optimized super-resolution algorithm was obtained through the multiscale network reconstruction model based on the traditional algorithm. A total of 100 patients undergoing craniotomy evacuation of hematoma were recruited and rolled into sevoflurane control group and propofol experimental group. Both were evaluated using diffusion tensor imaging (DTI) images based on deep learning super-resolution algorithms. The results showed that the fractional anisotropic image (FA) value of the hind limb corticospinal tract of the affected side of the internal capsule of the experimental group after the operation was 0.67 ± 0.28. The National Institute of Health Stroke Scale (NIHSS) score was 6.14 ± 3.29. The oxygen saturation in jugular venous (SjvO2) at T4 and T5 was 61.93 ± 6.58% and 59.38 ± 6.2%, respectively, and cerebral oxygen uptake rate (CO2ER) was 31.12 ± 6.07% and 35.83 ± 7.91%, respectively. The difference in jugular venous oxygen (Da-jvO2) at T3, T4, and T5 was 63.28 ± 10.15 mL/dL, 64.89 ± 13.11 mL/dL, and 66.03 ± 11.78 mL/dL, respectively. The neuron-specific enolase (NSE) and central-nerve-specific protein (S100β) levels at T5 were 53.85 ± 12.31 ng/mL and 7.49 ± 3.16 ng/mL, respectively. In terms of the number of postoperative complications, the patients in the experimental group were better than the control group under sevoflurane anesthesia, and the differences were substantial ( P  < 0.05). In conclusion, MRI images based on deep learning super-resolution algorithm have great clinical value in evaluating the degree of brain injury in patients anesthetized with propofol and the protective effect of propofol on brain nerves.

scholarly journals SMORE: A Self-supervised Anti-aliasing and Super-resolution Algorithm for MRI Using Deep Learning

2020 ◽  
pp. 1-1
Author(s):  
Can Zhao ◽  
Blake E. Dewey ◽  
Dzung L. Pham ◽  
Peter A. Calabresi ◽  
Daniel S. Reich ◽  
...  
Keyword(s):  
Deep Learning ◽  
Super Resolution ◽  
Resolution Algorithm

scholarly journals Deep‐learning based super‐resolution for 3D isotropic coronary MR angiography in less than a minute

2021 ◽  
Author(s):  
Thomas Küstner ◽  
Camila Munoz ◽  
Alina Psenicny ◽  
Aurelien Bustin ◽  
Niccolo Fuin ◽  
...  
Keyword(s):  
Deep Learning ◽  
Mr Angiography ◽  
Super Resolution ◽  
Coronary Mr Angiography

scholarly journals 29.2: Deep Learning Based Video Super Resolution A Survey

2021 ◽  
Vol 52 (S1) ◽  
pp. 187-187
Author(s):  
Yanpeng Cao ◽  
Feng Yu ◽  
Yongming Tang
Keyword(s):  
Deep Learning ◽  
Super Resolution

Efficient downscaling of satellite oceanographic data with convolutional neural networks

2021 ◽  
Vol 12 (3) ◽  
pp. 46-47
Author(s):  
Nikita Saxena
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Super Resolution ◽  
Sea Surface ◽  
Clear Sky ◽  
Oceanographic Data ◽  
Computationally Expensive ◽  
Sky Conditions ◽  
Data Inference

Space-borne satellite radiometers measure Sea Surface Temperature (SST), which is pivotal to studies of air-sea interactions and ocean features. Under clear sky conditions, high resolution measurements are obtainable. But under cloudy conditions, data analysis is constrained to the available low resolution measurements. We assess the efficiency of Deep Learning (DL) architectures, particularly Convolutional Neural Networks (CNN) to downscale oceanographic data from low spatial resolution (SR) to high SR. With a focus on SST Fields of Bay of Bengal, this study proves that Very Deep Super Resolution CNN can successfully reconstruct SST observations from 15 km SR to 5km SR, and 5km SR to 1km SR. This outcome calls attention to the significance of DL models explicitly trained for the reconstruction of high SR SST fields by using low SR data. Inference on DL models can act as a substitute to the existing computationally expensive downscaling technique: Dynamical Downsampling. The complete code is available on this Github Repository.

scholarly journals Real-Time Environment Monitoring Using a Lightweight Image Super-Resolution Network

2021 ◽  
Vol 18 (11) ◽  
pp. 5890
Author(s):  
Qiang Yu ◽  
Feiqiang Liu ◽  
Long Xiao ◽  
Zitao Liu ◽  
Xiaomin Yang
Keyword(s):  
Deep Learning ◽  
Real Time ◽  
Super Resolution ◽  
Model Complexity ◽  
Practical Application ◽  
Single Image ◽  
Feature Maps ◽  
Benchmark Datasets ◽  
Image Super Resolution ◽  
Single Image Super Resolution

Deep-learning (DL)-based methods are of growing importance in the field of single image super-resolution (SISR). The practical application of these DL-based models is a remaining problem due to the requirement of heavy computation and huge storage resources. The powerful feature maps of hidden layers in convolutional neural networks (CNN) help the model learn useful information. However, there exists redundancy among feature maps, which can be further exploited. To address these issues, this paper proposes a lightweight efficient feature generating network (EFGN) for SISR by constructing the efficient feature generating block (EFGB). Specifically, the EFGB can conduct plain operations on the original features to produce more feature maps with parameters slightly increasing. With the help of these extra feature maps, the network can extract more useful information from low resolution (LR) images to reconstruct the desired high resolution (HR) images. Experiments conducted on the benchmark datasets demonstrate that the proposed EFGN can outperform other deep-learning based methods in most cases and possess relatively lower model complexity. Additionally, the running time measurement indicates the feasibility of real-time monitoring.

Sign in / Sign up

Export Citation Format

Share Document