A5100348714- Medical Image Segmentation Techniques
- AI in cancer detection
- Advanced Neural Network Applications
- Radiomics and Machine Learning in Medical Imaging
- Mobile Crowdsensing and Crowdsourcing
- Medical Imaging Techniques and Applications
- Diverse Topics in Contemporary Research
- Face recognition and analysis
- Advanced X-ray and CT Imaging
- Open Source Software Innovations
Beijing Institute of Technology
2019-2024
Vanderbilt University
2024
Medical image segmentation is widely used in clinical diagnosis, and methods based on convolutional neural networks have been able to achieve high accuracy. However, it still difficult extract global context features, the parameters are too large be clinically applied. In this regard, we propose a novel network structure improve traditional encoder-decoder model, which saves while maintaining We feature extraction efficiency by constructing an encoder module that can simultaneously local...
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)...
With the characteristics of low cost and open call, crowdsourcing has been widely adopted in many fields, particularly to support use surveys, data processing, monitoring public health. The objective current study is analyze applications, hotspots, emerging trends field Using CiteSpace for visualization scientific maps, this explores analysis time-scope, countries institutions, authors, published journals, keywords, co-references, citation clusters. results show that United States country...
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....
The automatic medical image segmentation technology based on deep learning has achieved high accuracy, but there is a dilemma that the parameters are huge and difficult to clinical application. To solve this problem, we propose an Efficient Segmentation Network (ESNet) model saves floating point of operations while ensuring accuracy. We design powerful feature extraction module DME construct hierarchical encoding sub-network enable our have efficient capabilities, reuse-based DCD decoding...