A5086474293- Advanced MRI Techniques and Applications
- Medical Imaging Techniques and Applications
- Photoacoustic and Ultrasonic Imaging
- Advanced X-ray and CT Imaging
- Acute Myeloid Leukemia Research
- Image and Signal Denoising Methods
- Antioxidant Activity and Oxidative Stress
- Retinoids in leukemia and cellular processes
- Seismic Imaging and Inversion Techniques
- MRI in cancer diagnosis
- Image Processing Techniques and Applications
- Electron and X-Ray Spectroscopy Techniques
- Advanced Neuroimaging Techniques and Applications
- Medical Image Segmentation Techniques
- Cell Image Analysis Techniques
- Advanced Electron Microscopy Techniques and Applications
- Advanced Vision and Imaging
- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging and Analysis
- Atomic and Subatomic Physics Research
ShanghaiTech University
2022-2025
First Affiliated Hospital of Harbin Medical University
2019
Harbin Medical University
2019
Emerging unsupervised implicit neural representation (INR) methods, such as NeRP, NeAT, and SCOPE, have shown great potential to address sparse-view computed tomography (SVCT) inverse problems. Although these INR-based methods perform well in relatively dense SVCT reconstructions, they struggle achieve comparable performance supervised sparser scenarios. They are prone being affected by noise, limiting their applicability real clinical settings. Additionally, current not fully explored the...
Emerging unsupervised implicit neural representation (INR) methods, such as NeRP, NeAT, and SCOPE, have shown great potential in addressing sparse-view computed tomography (SVCT) inverse problems. While these INR-based methods perform well on relatively dense SVCT reconstructions, they struggle to achieve comparable performance with supervised sparser scenarios are prone being affected by noise, limiting their applicability real clinical settings. Additionally, current not fully explored the...
Limited-angle and sparse-view computed tomography (LACT SVCT) are crucial for expanding the scope of X-ray CT applications. However, they face challenges due to incomplete data acquisition, resulting in diverse artifacts reconstructed images. Emerging implicit neural representation (INR) techniques, such as NeRF, NeAT, NeRP, have shown promise under-determined imaging reconstruction tasks. unsupervised nature INR architecture imposes limited constraints on solution space, particularly highly...
High-resolution electron microscopy (HREM) imaging technique is a powerful tool for directly visualizing broad range of materials in real-space. However, it faces challenges denoising due to ultra-low signal-to-noise ratio (SNR) and scarce data availability. In this work, we propose Noise2SR, zero-shot self-supervised learning (ZS-SSL) framework HREM. Within our framework, super-resolution (SR) based training strategy, incorporating the Random Sub-sampler module. The designed generate...
Motion correction (MoCo) in radial MRI is a challenging problem due to the unpredictability of subject's motion. Current state-of-the-art (SOTA) MoCo algorithms often use extensive high-quality MR images pre-train neural networks, obtaining excellent reconstructions. However, need for large-scale datasets significantly increases costs and limits model generalization. In this work, we propose Moner, an unsupervised method that jointly solves artifact-free accurate motion from undersampled,...
Denoising of magnetic resonance image (MRI) is a critical step in MRI processing and analysis. With the advantage not requiring paired noisy-clean images for training, self-supervised denoising methods are emerging as competitive alternatives to supervised denoising. However, current effective enough MRI. In this work, we propose Noise2SR-M (N2SR-M), self method MR images, which more efficient high-dimensional images. N2SR-M designed training with noisy data different sizes divided from...
Magnetic Resonance Imaging (MRI) stands as a powerful modality in clinical diagnosis. However, it is known that MRI faces challenges such long acquisition time and vulnerability to motion-induced artifacts. Despite the success of many existing motion correction algorithms, there has been limited research focused on correcting artifacts estimated coil sensitivity maps for fast reconstruction. Existing methods might suffer from severe performance degradation due error propagation resulting...
Fluorescence microscopy is a key driver to promote discoveries of biomedical research. However, with the limitation microscope hardware and characteristics observed samples, fluorescence images are susceptible noise. Recently, few self-supervised deep learning (DL) denoising methods have been proposed. training efficiency performance existing relatively low in real scene noise removal. To address this issue, paper proposed image method Noise2SR (N2SR) train simple effective model based on...