A5052814363- Medical Image Segmentation Techniques
- Brain Tumor Detection and Classification
- Advanced Neural Network Applications
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Radiomics and Machine Learning in Medical Imaging
- Dementia and Cognitive Impairment Research
- AI in cancer detection
- Advanced Image Fusion Techniques
- Functional Brain Connectivity Studies
- Advanced MRI Techniques and Applications
- Image Processing Techniques and Applications
- Advanced Vision and Imaging
- Cardiovascular Function and Risk Factors
- Advanced Image Processing Techniques
- Neonatal and fetal brain pathology
- Colorectal Cancer Screening and Detection
- Pediatric Urology and Nephrology Studies
- Alzheimer's disease research and treatments
- Generative Adversarial Networks and Image Synthesis
- MRI in cancer diagnosis
- Remote-Sensing Image Classification
- Cardiovascular and exercise physiology
- Bone fractures and treatments
- Ferroptosis and cancer prognosis
Yantai University
2015-2024
Anhui Medical University
2022-2023
Sir Run Run Shaw Hospital
2023
Zhejiang University
2023
Anhui Provincial Center for Disease Control and Prevention
2022
Ministry of Education of the People's Republic of China
2022
University of Pennsylvania
2018-2020
Fudan University Shanghai Cancer Center
2020
Children's Hospital of Philadelphia
2020
Institute of Automation
2018
Abstract Individuals with mild cognitive impairment (MCI) of different subtypes show distinct alterations in network patterns. The first aim this study is to identify the MCI by employing a regional radiomics similarity (R2SN). second characterize abnormality patterns associated clinical manifestations each subtype. An individual‐level R2SN constructed for N = 605 normal controls (NCs), 766 patients, and 283 Alzheimer's disease (AD) patients. patients’ profiles are clustered into two using...
Background Radiographic measurement of leg length discrepancy (LLD) is time consuming yet cognitively simple for pediatric radiologists. Purpose To compare deep learning (DL) measurements LLD in patients to performed by Materials and Methods For this HIPAA-compliant retrospective study, radiographs obtained evaluate children between January August 2018 were identified. was automatically measured means image segmentation followed calculation. On training data, a DL model trained segment...
Abstract A structural covariance network (SCN) has been used successfully in magnetic resonance imaging (sMRI) studies. However, most SCNs have constructed by a unitary marker that is insensitive for discriminating different disease phases. The aim of this study was to devise novel regional radiomics similarity (R2SN) could provide more comprehensive information morphological analysis. R2SNs were computing the Pearson correlations between features extracted from any pair regions each...
Liver resection is the first-line treatment for primary liver cancers, providing potential a cure. However, concerns about post-hepatectomy failure (PHLF), leading cause of death following extended resection, have restricted population eligible patients. Here, we engineered clinical-grade bioartificial (BAL) device employing human-induced hepatocytes (hiHeps) manufactured under GMP conditions. In porcine PHLF model, hiHep-BAL showed remarkable survival benefit. On top supportive function,...
Digital histopathology image segmentation can facilitate computer-assisted cancer diagnostics. Given the difficulty of obtaining manual annotations, weak supervision is more suitable for task than full is. However, most weakly supervised models are not ideal handling severe intra-class heterogeneity and inter-class homogeneity in images. Therefore, we propose a novel end-to-end learning framework named WESUP. With only sparse point it performs accurate exhibits good generalizability. The...
Classification of ultrasound (US) kidney images for diagnosis congenital abnormalities the and urinary tract (CAKUT) in children is a challenging task. It desirable to improve existing pattern classification models that are built upon conventional image features. In this study, we propose transfer learning-based method extract imaging features from US order CAKUT children. Particularly, pre-trained deep learning model (imagenet-caffe-alex) adopted feature extraction 3-channel maps computed...
We introduce MindSpore Quantum, a pioneering hybrid quantum-classical framework with primary focus on the design and implementation of noisy intermediate-scale quantum (NISQ) algorithms. Leveraging robust support MindSpore, an advanced open-source deep learning training/inference framework, Quantum exhibits exceptional efficiency in training variational algorithms both CPU GPU platforms, delivering remarkable performance. Furthermore, this places strong emphasis enhancing operational when...
A novel label fusion method for multi-atlas based image segmentation is developed by integrating semi-supervised and supervised machine learning techniques. Particularly, our in a pattern recognition framework. We build random forests classification models each voxel to be segmented on its corresponding patches of atlas images that have been registered the segmented. The voxelwise are then applied obtain probabilistic map. Finally, propagation adapted refine map propagating reliable labels,...
Background: Type 2 diabetes mellitus (T2DM) is a common risk factor for cardiovascular diseases. The aims of this study were to evaluate the changes in left ventricular myocardial work T2DM patients using pressure-strain loop (PSL) technique, and explore factors impairment. Methods: Fifty with 50 normal controls (NCs) included study. In addition conventional echocardiography two-dimensional speckle tracking echocardiography, parameters measured PSL technology. Results: absolute value global...
Abstract Many unsupervised methods are widely used for parcellating the brain. However, aren’t able to integrate prior information, obtained from such as exiting functional neuroanatomy studies, parcellate brain, whereas information guided semi-supervised method can generate more reliable brain parcellation. In this study, we propose a novel clustering into spatially and functionally consistent parcels based on resting state magnetic resonance imaging (fMRI) data. Particularly, supervised...
Band selection is an effective means to alleviate the curse of dimensionality in hyperspectral data. Many methods select a compact and low redundant band subset, which inadequate as it may degrade classification performance. Instead, more emphasis shall be put on selecting representative bands. In this article, we propose robust unsupervised method address issue. Our reveals bandwise representativeness based comprehensive interband neighborhood structure. It incorporates graph into sparse...
Background Dynamic contrast material-enhanced MR lymphangiography has recently emerged as a technique to image the lymphatic anatomy and identify flow abnormalities; however, method quantify in health disease is needed. Purpose To develop thoracic patterns using dynamic contrast-enhanced lymphangiography. Materials Methods The following patients with images collected 2015 2016 were retrospectively identified: group A, neonates chylothorax; B, children heart failure complicated by plastic...
A remote sensing image (RSI) fusion method based on multiscale morphological component analysis (m-MCA) is presented. Our contribution describes a new sparse decomposition algorithm called m-MCA, which we apply to RSI fusion. Building MCA, m-MCA combines curvelet transform bases and local discrete cosine build dictionary, controls the entries of dictionary decompose into texture components cartoon with different scales. The effective scale high-resolution multispectral are selected...