A5100644288- Radiomics and Machine Learning in Medical Imaging
- Gene expression and cancer classification
- Medical Image Segmentation Techniques
- MRI in cancer diagnosis
- Iron and Steelmaking Processes
- Machine Learning in Bioinformatics
- Evolutionary Algorithms and Applications
- Bioinformatics and Genomic Networks
- Advanced X-ray and CT Imaging
- MicroRNA in disease regulation
- Thermochemical Biomass Conversion Processes
- Topic Modeling
- Artificial Intelligence in Healthcare
- Law, AI, and Intellectual Property
- Innovations in Concrete and Construction Materials
- Medical Research and Treatments
- Genetic and phenotypic traits in livestock
- Single-cell and spatial transcriptomics
- Pancreatic and Hepatic Oncology Research
- Face and Expression Recognition
- Chronic Kidney Disease and Diabetes
- Optical Coherence Tomography Applications
- Cleft Lip and Palate Research
- Machine Learning in Healthcare
- Autopsy Techniques and Outcomes
Northeastern University
2011-2024
Ministry of Education of the People's Republic of China
2013-2022
University of Chicago
2020
NorthShore University HealthSystem
2020
Shenyang Institute of Computing Technology (China)
2020
Affiliated Hospital of Qingdao University
2017
Qingdao University
2017
82th Hospital of Pla
2013
Huadong Hospital
2012
Fudan University
2012
Breast cancer is a major cause of mortality among women if not treated in early stages. Recognizing molecular markers from DCE-MRI directly to distinguish the four subtypes without invasive biopsy helpful for guiding treatment plans breast cancer, which provides fast way consequential plan decision time and best opportunity patients. This study presents an approach recognition image phenotypes by radiomics. An improved region growth algorithm with dynamic threshold user interaction proposed...
Background: The estimated glomerular filtration rate (eGFR) is frequently used to monitor progression of kidney disease. Multiple values have be obtained, sometimes over years determine the decline in function. Recent data suggest that functional MRI (fMRI) methods may able predict loss eGFR. In a prior study, baseline with multi-parametric individuals diabetes and moderate CKD was reported. This report extends our observations order evaluate temporal variability fMRI measurements 36 months...
In bioinformatics, the rapid development of gene sequencing technology has produced an increasing amount microarray data. This type data shares typical characteristics small sample size and high feature dimensions. Searching for biomarkers from data, which expression features various diseases, is essential disease classification. selection therefore became fundemental analysis designs to remove irrelevant redundant features. There are a large number in severely degrade classification...
Image segmentation is still an open problem especially when intensities of the objects interest are overlapped due to presence intensity inhomogeneities. A bias correction embedded level set model proposed in this paper where inhomogeneities estimated by orthogonal primary functions. First, inhomogeneous clustering energy defined based on global distribution characteristics image intensities, and membership functions clusters described function then introduced define data term model. Second,...
The development of medical device technology has led to the rapid growth imaging data. reconstruction from two-dimensional images three-dimensional volume visualization not only shows location and shape lesions multiple views but also provides intuitive simulation for surgical treatment. However, process requires high performance execution image data acquisition algorithms, which limits application equipments with limited resources. Therefore, it is difficult apply on many online scenarios,...
In the field of genomics and personalized medicine, it is a key issue to find biomarkers directly related diagnosis specific diseases from high-throughput gene microarray data. Feature selection technology can discover with disease classification information.We use support vector machines as classifiers five-fold cross-validation average accuracy, recall, precision F1 score evaluation metrics evaluate identified biomarkers. Experimental results show accuracy above 0.93, recall 0.92, 0.91,...
<abstract> <p>Gene expression data is highly dimensional. As disease-related genes account for only a tiny fraction, deep learning model, namely GSEnet, proposed to extract instructive features from gene data. This model consists of three modules, the pre-conv module, SE-Resnet and SE-conv module. Effectiveness on performance improvement 9 representative classifiers evaluated. Seven evaluation metrics are used this assessment GSE99095 dataset. Robustness advantages compared with...
The rapid development of microarray technology has generated a large amount data, and the classification these data is meaningful for cancer diagnosis, treatment prognosis. high-dimensional with small samples challenging problem, which usually requires feature selection methods to reduce dimensionality first. However, different generate lists same data. Researchers need choose among many methods, reduces research efficiency. Therefore, rank aggregation method used optimal list by aggregating...
Medical activities recommendation is a key aspect of an intelligent healthcare system, which can assist doctors with little clinical experience in decision making. be seen as kind temporal set prediction. Previous studies about them are based on Recurrent Neural Network (RNN), does not incorporate personalized medical history or differentiate between the impact activities. To address above-given issues, this paper proposes Next-Day Activities Recommendation (NDMARec) model. Specifically, our...
Segmentation of the femoral head in CT images is a very important and challenging problem medical image analysis. This paper proposes fully automatic segmentation method based on graph cuts model shape constraint. Firstly, we use Three-dimensional Otsu Thresholding to pre-segment acetabular from images. Then choose optimize results for subsequent detection. Next, find position circle where located extract new features according this detected separate acetabulum. Finally, automated achieved.
Diabetic retinopathy (DR) and age-related macular degeneration (AMD) are important causes of blindness visual loss. Optical coherence tomography (OCT) is a non-invasive optical imaging method that can capture retinal vascular information even pathological information. In order to improve the screening rate accuracy these two diseases, we propose new network structure named TCAM-Resnet, which uses OCT three-dimensional images screen classify AMD DR. TCAM-Resnet based on Resnet network. A...
The liquid-gas two-phase flow in rock fractures is of great significance the community civil engineering. velocity a fluid through fracture depends on both hydraulic conductivity and applied pressure drop along fracture. capillary intermediates by altering "effective" acting gas liquid phases flowing fractures. This paper presents model, into which influence was introduced to quantify interaction between phase phase. found exhibit different patterns at rates, i.e., bubble flow, slug annular...
The sequence bio-film reactor and traditional biological aerated filter system, investigate the sewage treatment under 2/3 volume running performance, compare with all running. result shows that is good of SBBR-B AF. average removal rates are 92.70%, 92.19%, 83.05% 95.28% effluent concentration 17.96 mg/L, 2.05mg/L, 5.7mg/L 0.32 mg/L. Satisfy -B criteria specified in Discharge Standard Pollutants for Municipal Wastewater Treatment Plant (GB 18918-2002). Denitrify phosphorus close running,...