A5100378393- Radiomics and Machine Learning in Medical Imaging
- COVID-19 diagnosis using AI
- Advanced Neural Network Applications
- AI in cancer detection
- Electricity Theft Detection Techniques
- Smart Grid and Power Systems
- Medical Image Segmentation Techniques
- Spectroscopy and Chemometric Analyses
- Domain Adaptation and Few-Shot Learning
- Medical Imaging and Analysis
- Brain Tumor Detection and Classification
- Cardiovascular Function and Risk Factors
- Ocular and Laser Science Research
- Spectroscopy Techniques in Biomedical and Chemical Research
- Smart Grid Energy Management
- Multimodal Machine Learning Applications
- Topic Modeling
- Digital Imaging for Blood Diseases
- Cardiac Imaging and Diagnostics
- Lung Cancer Diagnosis and Treatment
- Acute Ischemic Stroke Management
Shanghai Center for Brain Science and Brain-Inspired Technology
2023-2024
Shandong University
2012-2024
University of Jinan
2022-2023
Shandong Youth University of Political Science
2022
Zhongyuan University of Technology
2022
The University of Texas Southwestern Medical Center
2022
Wenzhou Medical University
2020
Background: On December 8, 2019, the first new coronavirus case was discovered in Wuhan, China, and an intensive outbreak incepted next month (about January 20). Virologicalists epidemiologists predict that it will reach a peak about 90 days fade away till end 4 months (Early April), entire epidemic terminate early May. The daily rise confirmed cases increase number of communities on map always hit nerves panic. 31, 2020, World Health Organization (WHO) declared China 's "public health...
Ischemic stroke is the most common brain disease. Segmentation of lesion from a medical scan vital to plan surgical procedure. The purpose this work develop an efficient network based on self-attention and spatial-channel attention mechanisms. In letter, novel multi-encoder transformer (METrans) proposed, which overcomes inability U-Nets model long-range contextual interactions. Different traditional segmentation methods, four encoders with different scales are explored extract multi-scale...
In clinical settings, the implementation of deep neural networks is impeded by prevalent problems label scarcity and class imbalance in medical images. To mitigate need for labeled data, semi-supervised learning (SSL) has gained traction. However, existing SSL schemes exhibit certain limitations. 1) They commonly fail to address problem. Training with imbalanced data makes model's prediction biased towards majority classes, consequently introducing bias. 2) usually suffer from training bias...
Accurate identification of lesions is a key step in surgical planning. However, this task mainly exists two challenges: 1) Due to the complex anatomical shapes different lesions, most segmentation methods only achieve outstanding performance for specific structure, rather than other with location differences. 2) The huge number parameters limits existing transformer-based models. To overcome these problems, we propose novel slight dual-path network (SDPN) segment variable or organs...
Accurate and fully automated brain structure examination prediction from 3D volumetric magnetic resonance imaging (MRI) is a necessary step in medical analysis, which can assist greatly clinical diagnosis. Traditional deep learning models suffer severe performance degradation when applied to clinically acquired unlabeled data. The mainly caused by domain discrepancy such as different device types parameter settings for data acquisition. However, existing approaches focus on the reduction of...
Corona Virus Disease 2019 (COVID-19) spread globally in early 2020, leading to a new health crisis. Automatic segmentation of lung infections from computed tomography (CT) images provides an important basis for diagnosis COVID-19 quickly. In this paper, we propose effective Lung Infection Segmentation Network (LISNet) based on edge supervision and multi-scale context aggregation. More specifically, Edge Supervision module is introduced the feature extraction part enhance low contrast between...
Objective. It was a great challenge to train an excellent and generalized model on ultra-small data set composed of multi-orientation cardiac cine magnetic resonance imaging (MRI) images. We try develop 3D deep learning method based training from muti-orientation MRI images assess its performance automated biventricular structure segmentation function assessment in multivendor.Approach. completed the testing our networks using only heart datasets 150 cases (90 for 60 testing). This were...
The conventional pretraining-and-finetuning paradigm, while effective for common diseases with ample data, faces challenges in diagnosing data-scarce occupational like pneumoconiosis. Recently, large language models (LLMs) have exhibits unprecedented ability when conducting multiple tasks dialogue, bringing opportunities to diagnosis. A strategy might involve using adapter layers vision-language alignment and diagnosis a dialogic manner. Yet, this approach often requires optimization of...