A5068127687- Advanced Electron Microscopy Techniques and Applications
- AI in cancer detection
- Electron and X-Ray Spectroscopy Techniques
- Nanopore and Nanochannel Transport Studies
- COVID-19 diagnosis using AI
- Digital Imaging for Blood Diseases
- Digital Holography and Microscopy
- Radiomics and Machine Learning in Medical Imaging
- RNA and protein synthesis mechanisms
- Advanced Neural Network Applications
- Machine Learning in Healthcare
- Geological and Geophysical Studies
- Force Microscopy Techniques and Applications
- Advanced Image Processing Techniques
- Image Retrieval and Classification Techniques
- Advanced X-ray Imaging Techniques
- RNA modifications and cancer
- Cell Image Analysis Techniques
- Generative Adversarial Networks and Image Synthesis
- Colorectal Cancer Surgical Treatments
- Emergency and Acute Care Studies
- Brain Tumor Detection and Classification
- ATP Synthase and ATPases Research
- Computational Physics and Python Applications
- Colorectal Cancer Screening and Detection
Tsinghua University
2016-2025
Tencent (China)
2019-2024
Center for Life Sciences
2017-2024
Abstract The sudden deterioration of patients with novel coronavirus disease 2019 (COVID-19) into critical illness is major concern. It imperative to identify these early. We show that a deep learning-based survival model can predict the risk COVID-19 developing based on clinical characteristics at admission. develop this using cohort 1590 from 575 medical centers, internal validation performance concordance index 0.894 further validate three separate cohorts Wuhan, Hubei and Guangdong...
Multiple instance learning (MIL) is a typical weakly-supervised method where the label associated with bag of instances instead single instance. Despite extensive research over past years, effectively deploying MIL remains an open and challenging problem, especially when commonly assumed standard multiple (SMI) assumption not satisfied. In this paper, we propose based on deep graph convolutional network feature selection (FS-GCN-MIL) for histopathological image classification. The proposed...
Ptychography is an enabling coherent diffraction imaging technique for both fundamental and applied sciences. Its applications in optical microscopy, however, fall short its low throughput limited resolution. Here, we report a resolution-enhanced parallel coded ptychography achieving the highest numerical aperture orders of magnitude greater than previous demonstrations. In this platform, translate samples across disorder-engineered surfaces lensless data acquisition. The engineered surface...
Mitotic cell detection and counting per tissue area is an important aggressiveness indicator for the invasive breast cancer. However, manual mitosis by pathologists extremely labor-intensive. Several automatic methods have been proposed in recent years. Traditional using hand-crafted features suffer from large mitotic shape variation existence of many mimics with similar appearance. Pixel-wise classification working a sliding window manner time-consuming which hinders it clinical...
The resolution of electron-potential maps in single-particle cryo-electron microscopy (cryoEM) is approaching atomic or near- resolution. However, no program currently exists for de novo cryoEM model building at resolutions exceeding beyond 3.5 Å. Here, we present a program, EMBuilder, based on template matching, to generate models high identifies features both secondary-structure and Cα stages. In the secondary structure stage, helices strands are identified with pre-computed templates,...
Abstract Advancements in cryo-electron tomography (cryoET) allow the structure of macromolecules to be determined situ, which is crucial for studying membrane protein structures and their interactions cellular environment. However, membranes are often highly curved have a strong contrast cryoET tomograms, masks signals from proteins. These factors pose difficulties observing revealing proteins situ. Here, we report membrane-flattening method corresponding software, MPicker, designed...
This study aimed to develop an artificial intelligence model for predicting the pathological complete response (pCR) neoadjuvant chemoradiotherapy (nCRT) of locally advanced rectal cancer (LARC) using digital images.nCRT followed by total mesorectal excision (TME) is a standard treatment strategy patients with LARC. Predicting PCR nCRT LARC remine difficulty.842 treated from three medical centers were retrospectively recruited and subgrouped into training, testing external validation sets....
<title>Abstract</title> Advancements in cryo-electron tomography (cryoET) allow the structure of macromolecules to be determined <italic>in situ</italic>, which is crucial for studying membrane protein structures and their interactions cellular environment. However, membranes are often highly curved have a strong contrast cryoET tomograms, masks signals from proteins. These factors pose difficulties observing revealing proteins situ</italic>. Here, we report membrane-flattening method...
As single particle cryo-electron microscopy has evolved to a new era of atomic resolution, sample heterogeneity still imposes major limit the resolution many macromolecular complexes, especially those with continuous conformational flexibility. Here, we describe segmentation algorithm towards solving structures molecules composed several parts that are relatively flexible each other. In this algorithm, different target molecule segmented from raw images according their alignment information...
Cryo-EM in single particle analysis is known to have low SNR and requires utilize several frames of the same sample restore one high-quality image for visualizing that particle. However, cryo-EM movie motion caused by beam striking make task very challenging. Video enhancement algorithms computer vision shed new light on tackling such tasks utilizing deep neural networks. they are designed natural images with high SNR. Meanwhile, lack ground truth seems be major limiting factor progress....
Cryo-EM movie in single particle analysis has extremely low SNR and requires aligning multiple frames to achieve signal enhancement. Currently, processing technique is adopted estimate the motion vector between a pair of cryo-EM at patch-level, estimated used as reference for frame alignment, whose accuracy will determine resolution reconstructed 3D structure particle. The patch-level may not well represent beam-induced particles since patch move towards different directions due beam...
Metastatic involvement of lymph nodes is one the most important prognostic variables for many cancers. Several deep learning based algorithms have been developed to segment metastatic regions in pathological images help predict prognosis. However, training these methods requires a large amount annotated data, and labeling task an extremely time-consuming process human annotators. In order reduce annotation burden, we first time propose semi-automatic method (PiPo-Net) images. The comprised...