A5101521563- Surgical Simulation and Training
- Advanced Neural Network Applications
- Functional Brain Connectivity Studies
- Neural dynamics and brain function
- Climate Change and Health Impacts
- AI in cancer detection
- Machine Learning and Data Classification
- Mental Health Research Topics
- EEG and Brain-Computer Interfaces
- COVID-19 and Mental Health
- Retinal Diseases and Treatments
- Anatomy and Medical Technology
- Metabolomics and Mass Spectrometry Studies
- Face recognition and analysis
- Digital Mental Health Interventions
- Domain Adaptation and Few-Shot Learning
- Machine Learning and Algorithms
- Urban Green Space and Health
- Health disparities and outcomes
- Advanced MRI Techniques and Applications
- Air Quality and Health Impacts
- Medical Imaging and Analysis
- Cell Adhesion Molecules Research
- Medical Image Segmentation Techniques
- E-commerce and Technology Innovations
University of Electronic Science and Technology of China
2022-2024
Indiana University Bloomington
2022-2024
Vanderbilt University
2023-2024
Xi'an Polytechnic University
2024
Chengdu University of Traditional Chinese Medicine
2023
Inner Mongolia People's Hospital
2022
Baoding People's Hospital
2022
Baoding No.1 Central Hospital
2022
China Medical University
2020
Changsha University of Science and Technology
2017
A common problem with segmentation of medical images using neural networks is the difficulty to obtain a significant number pixel-level annotated data for training. To address this issue, we proposed semi-supervised network based on contrastive learning. In contrast previous state-of-the-art, introduce Min-Max Similarity (MMS), learning form dual-view training by employing classifiers and projectors build all-negative, positive negative feature pairs, respectively, formulate as solving MMS...
The Segment Anything Model (SAM) is a recently developed all-range foundation model for image segmentation. It can use sparse manual prompts such as bounding boxes to generate pixel-level segmentation in natural images but struggles medical low-contrast, noisy ultrasound images. We propose refined test-phase prompt augmentation technique designed improve SAM's performance method couples multi-box and an aleatoric uncertainty-based false-negative (FN) false-positive (FP) correction (FNPC)...
We present SOmicsFusion, a software toolbox for 'fusing' spatial omics with classical biomedical imaging modalities, capitalizing on their inherent correspondences and complementarity when characterizing the same subject. By augmenting radiological histological images spatially resolved molecular profiling, this fusion offers panoramic characterization of biochemical perturbations underlying pathological conditions, thereby advancing our understanding diseases like brain disorders cancers....
Abstract Background Cocaine Use Disorder (CUD) poses significant neurobiological and neuropsychiatric challenges, often resulting in severe cognitive behavioral impairments. This study aims to explore the neural dynamics of CUD using a dynamic coactivation pattern (CAP) analysis approach provide deeper understanding transient mechanisms disorder. Methods Resting-state functional MRI data (SUDMEX_CONN) from 56 patients 57 healthy controls (HC) were analyzed. CAP was employed capture brain...
The Segment Anything Model (SAM) is a recently developed all-range foundation model for image segmentation. It can use sparse manual prompts such as bounding boxes to generate pixel-level segmentation in natural images but struggles medical low-contrast, noisy ultrasound images. We propose refined test-phase prompt augmentation technique designed improve SAM's performance method couples multi-box and an aleatoric uncertainty-based false-negative (FN) false-positive (FP) correction (FNPC)...
The ability to automatically detect and track surgical instruments in endoscopic videos can enable transformational interventions. Assessing performance efficiency, identifying skilled tool use choreography, planning operational logistical aspects of OR resources are just a few the applications that could benefit. Unfortunately, obtaining annotations needed train machine learning models identify localize tools is difficult task. Annotating bounding boxes frame-by-frame tedious...
Objective: To explore the clinical characteristics and recurrence rate of ischemic stroke in patients with antineutrophil cytoplasmic antibody (ANCA) , to improve early diagnosis these patients, so as take effective measures control progress disease reduce mortality. Methods: A retrospective analysis ANCA positive negative status from neurology department, Baoding No. 1 Central Hospital, Hebei, China. Results: The prevalence 1,297 patient was 10.10%. group significantly higher than that...
Abstract Despite the importance of daily stress to individuals' health and wellbeing, few studies have explored where happens in real time, that is, dynamic processes different spaces. As such, interventions rarely account for environment which occurs. We used mobile phone based ecological momentary assessment (EMA) collect data. Thirty‐three participants utilized a mobile‐phone‐based EMA app self‐report stressors as they went about their lives. Geographic coordinates were automatically...
Convolutional Neural Networks (CNNs) exhibit strong performance in medical image segmentation tasks by capturing high-level (local) information, such as edges and textures. However, due to the limited field of view convolution kernels, it is hard for CNNs fully represent global information. Recently, transformers have shown good their ability better model long-range dependencies. Nevertheless, struggle capture spatial features effectively CNNs. A should learn a representation from local be...
In mainland China, sexuality education is inadequate due to prevailing conservative social and cultural attitudes. This paper explores the role of online forums as alternative spaces for informal education, focusing on Douban group named "Gender Identity & Sexual Orientation Self-Study" (GISOS). Douban.com, a popular networking site in allows users create groups discussions various themes. To explore whether how GISOS can serve sources young people this study adopts qualitative research...
Ureteroscopic intrarenal surgery comprises the passage of a flexible ureteroscope through ureter into kidney and is commonly used for treatment stones or upper tract urothelial carcinoma (UTUC). Flexible ureteroscopes (fURS) are limited by their visualization ability fragility, which can cause missed regions during procedure in hard-to-visualize locations and/or due to scope breakage. This contributes high recurrence rate both stone UTUC patients. We introduce an automated patient-specific...
OBJECTIVES/GOALS: In this study, we implemented deformable medial modeling as a morphometric approach in first trimester placentas to characterize differences between fully automated and manual segmentations. METHODS/STUDY POPULATION: Twenty from singleton pregnancies 11-14 weeks’ gestation were manually automatically segmented 3D ultrasound volumes. Automated segmentations produced by trained convolutional neural network pipeline. Dice overlap scores volumes computed Deformable was applied...
Purpose To investigate how passive hyperthermia affect the resting-state functional brain activity based on an acute mouse model after heat stress exposure.
Placenta volume measurement from 3D ultrasound images is critical for predicting pregnancy outcomes, and manual annotation the gold standard. However, such expensive time-consuming. Automated segmentation algorithms can often successfully segment placenta, but these methods may not consistently produce robust segmentations suitable practical use. Recently, inspired by Segment Anything Model (SAM), deep learning-based interactive models have been widely applied in medical imaging domain....
Placenta volume measured from 3D ultrasound (3DUS) images is an important tool for tracking the growth trajectory and associated with pregnancy outcomes. Manual segmentation gold standard, but it time-consuming subjective. Although fully automated deep learning algorithms perform well, they do not always yield high-quality results each case. Interactive models could address this issue. However, there limited work on interactive placenta. Despite their accuracy, these methods may be feasible...
Ultrasound (US) image stitching can expand the field-of-view (FOV) by combining multiple US images from varied probe positions. However, registering with only partially overlapping anatomical contents is a challenging task. In this work, we introduce SynStitch, self-supervised framework designed for 2DUS stitching. SynStitch consists of synthetic pair generation module (SSPGM) and an (ISM). SSPGM utilizes patch-conditioned ControlNet to generate realistic pairs known affine matrix single...
Motivation: The effects of passive hyperthermia on brain function in resting-state mice are unknown. Goal(s): To expiore the rs-fMRI treated and normal brain. Approach: ICA network ROI-ROI FC were compared with control. Results: networks changed overall decreased. Impact: After prolonged exposure to high temperature has a greater impact perception cognitive level mice, which might help understand relationship between neuronal activities physiological thermal sensation regulation as well...
The paper is to reveal influential factors and the marginal role of urban residential land price in order provide guidance for making policies scientific use by GWR (geographically weighted regression) model, then displays spatial distribution GIS(geographic information system) drawing tools. As shown research results, metro sites, commercial outlets, water greenbelt-landscape has a conspicuous feature non-stationarity area, while other factor have an inconspicuous non-stationarity. FAR...