Article
Engineering, Biomedical
Ling Luo, Dingyu Xue, Feng Pan, Xinglong Feng
Summary: This study proposes a novel segmentation architecture, BGA-Net, which achieves state-of-the-art OD and OC segmentation performance on three publicly available datasets, highlighting the importance of calculated values in glaucoma assessment.
INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY
(2021)
Article
Computer Science, Artificial Intelligence
Baoliang Zhang, Xiaoxin Guo, Guangyu Li, Zhengran Shen, Xiaoying Hu, Songtian Che
Summary: In this paper, a novel one-stage framework called the multiple graph reasoning network (MGRNet) is proposed to segment optic discs (ODs) and optic cups (OCs) from different fundus image datasets. The MGRNet performs graph inference in three different spaces and achieves satisfactory segmentation results. It also provides reliable glaucoma screening results based on cup-to-disc ratio (CDR) measurement.
APPLIED INTELLIGENCE
(2023)
Article
Biology
Zhenzhong Liu, Laiwang Zheng, Lin Gu, Shubin Yang, Zichen Zhong, Guobin Zhang
Summary: In robot-assisted surgery, precise surgical instrument segmentation technology plays a crucial role in facilitating efficient and safe surgical operations. This article introduces an effective surgical instrument segmentation network called InstrumentNet, which utilizes YOLOv7 as the object detection framework to achieve real-time detection. Experimental results demonstrate that the proposed model achieves excellent segmentation performance on surgical instruments compared to other advanced models, highlighting its universality and superiority.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Computer Science, Artificial Intelligence
Pengshuai Yin, Yanwu Xu, Jinhui Zhu, Jiang Liu, Chang'an Yi, Huichou Huang, Qingyao Wu
Summary: The study introduces a level set based deep learning method for optic disc and cup segmentation, addressing the challenge of injecting domain-specific knowledge into existing segmentation networks. By adding constraints and considering pixel relationships, the proposed method effectively solves the problem. Experimental results confirm the effectiveness of the approach.
Article
Biology
Kaiwen Hua, Xianjin Fang, Zhiri Tang, Ying Cheng, Zekuan Yu
Summary: This paper proposes a novel framework called DCAM-NET for fundus domain generalization segmentation, which improves the model's ability to generalize to target domain data. It also introduces a multi-scale attention mechanism module (MSA) and a multi-region weight fusion convolution module (MWFC) to enhance the segmentation ability of the model on unknown domain data.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Mathematics
Zhijie Liu, Yuanqiong Chen, Xiaohua Xiang, Zhan Li, Bolin Liao, Jianfeng Li
Summary: This study proposes a lightweight segmentation algorithm, GlauNet, based on convolutional neural networks, for accurate segmentation of the optic disc and optic cup in the diagnosis of glaucoma. The algorithm achieves high segmentation efficiency and accuracy through an efficient feature-extraction network and a multiscale boundary fusion module.
Article
Multidisciplinary Sciences
Xiaoyu Tang, Sirui Liu, Qiuchi Xiang, Jintao Cheng, Huifang He, Bohuan Xue
Summary: Achieving emotion recognition in human-computer interaction is crucial in the era of artificial intelligence. We proposed a dual-channel network based on the Canny edge detector to improve facial expression recognition performance without adding redundant layers or training. Ablation experiments were conducted to discuss fusion parameters, and the method achieved good results in multiple datasets.
Article
Computer Science, Artificial Intelligence
Samiksha Pachade, Prasanna Porwal, Manesh Kokare, Luca Giancardo, Fabrice Meriaudeau
Summary: This paper presents a novel segmentation network called NENet, which combines deep learning frameworks with specific network structures to achieve superior performance and generalizability in optic disc and cup segmentation for glaucoma. Experimental results demonstrate that NENet outperforms current state-of-the-art methods in accuracy and detail segmentation.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Computer Science, Artificial Intelligence
Xin Yuan, Lingxiao Zhou, Shuyang Yu, Miao Li, Xiang Wang, Xiujuan Zheng
Summary: A new method proposed in this study successfully reduces the semantic gaps in the fusion of deep and shallow semantic information by using a residual multi-scale convolutional neural network with a context semantic extraction module to jointly segment the optic disc and optic cup in color fundus images.
ARTIFICIAL INTELLIGENCE IN MEDICINE
(2021)
Article
Engineering, Multidisciplinary
Qitong Chen, Liang Chen, Qi Li, Juanjuan Shi, Zhongkui Zhu, Changqing Shen
Summary: This article proposes a lightweight and robust model for cross-domain bearing fault diagnosis. By using feature fusion-based unsupervised adversarial learning, the model addresses the weaknesses of large size, complex calculation, and weak anti-noise ability. Experimental results demonstrate that the proposed model outperforms existing methods in terms of size, computation, and robustness.
Article
Computer Science, Artificial Intelligence
Amin Golzari Oskouei, Mahdi Hashemzadeh, Bahareh Asheghi, Mohammad Ali Balafar
Summary: The FCM algorithm is a popular method for data clustering and image segmentation but it is sensitive to the initialization of primary clusters and has equal importance of image features. This paper proposes the CGFFCM method, which improves segmentation performance through automatic weighting and feature strategies, and outperforms competitors when evaluated using the Berkeley benchmark dataset.
APPLIED SOFT COMPUTING
(2021)
Article
Environmental Sciences
Lei Huang, Min Li, Tao Xu, Shao-Qun Dong
Summary: Garbage recycling and automatic sorting are efficient ways to address the paradox of rising municipal waste. In this paper, we propose a trash picture categorization model called ResMsCapsule network, which combines the residual network and multi-scale module to greatly improve the performance of the basic capsule network. Extensive experiments using the TrashNet dataset show that the ResMsCapsule method has a simpler network structure and higher garbage classification accuracy than other image classification algorithms. The ResMsCapsule network achieves a classification accuracy of 91.41% with only 40% of the parameters of ResNet18.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Environmental Sciences
Kai Hu, Meng Li, Min Xia, Haifeng Lin
Summary: This study proposes a multi-scale feature aggregation network for water area segmentation. By designing a deep feature extraction module and a multi-branch aggregation module, it accurately identifies small tributaries in water area images and extracts deep semantic information, achieving improved segmentation accuracy.
Article
Engineering, Biomedical
Lei Yang, Yuge Gu, Guibin Bian, Yanhong Liu
Summary: This paper proposes a multi-scale attention fusion network (MAF-Net) to address the limitations of existing segmentation networks in processing micro objects and local semantic features, thereby improving the accuracy of surgical instrument segmentation.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Biology
Zhaojin Fu, Jinjiang Li, Zhen Hua
Summary: In this paper, we propose a feature attention network based on dual encoder, which achieves excellent performance in learning and extracting detailed features. It has important applications in the field of medical image segmentation.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Materials Science, Textiles
Jun Wu, Lin Wang, Zhitao Xiao, Lei Geng, Fang Zhang, Yanbei Liu
Summary: In this study, a double-branch deep cross-level fusion convolutional neural network (D-DCFNet) was proposed to improve feature selection in the objective evaluation of wool knitted fabric pilling. Experimental results showed that D-DCFNet achieved high accuracy rates for woolen and semi-worsted knitted fabrics with a small model size.
JOURNAL OF THE TEXTILE INSTITUTE
(2021)
Article
Computer Science, Artificial Intelligence
Jun Wu, Juan Le, Zhitao Xiao, Fang Zhang, Lei Geng, Yanbei Liu, Wen Wang
Summary: This paper proposes a wide-and-light network structure based on Faster R-CNN to improve the accuracy and reduce the computational cost of fabric defect detection. By designing dilated convolution modules and optimizing feature extraction networks, high-precision fabric defect detection is achieved.
APPLIED INTELLIGENCE
(2021)
Article
Engineering, Electrical & Electronic
Jun Wu, Qianqian Zhang, Mengjia Liu, Zhitao Xiao, Fang Zhang, Lei Geng, Yanbei Liu, Wen Wang
Summary: Early detection and accurate grading of DME is crucial for reducing the risk of vision loss in diabetic patients. The proposed DME grading method showed high accuracy in testing, improving efficiency and saving medical resources compared to other commonly used methods.
SIGNAL IMAGE AND VIDEO PROCESSING
(2021)
Article
Engineering, Electrical & Electronic
Fang Zhang, Dongxu Zhao, Zhitao Xiao, Jun Wu, Lei Geng, Wen Wang, Yanbei Liu
Summary: This paper presents an automated procedure for expedited parameter measurement of rodlike nanoparticles, including key steps such as nanoparticle segmentation. By using the Mask R-CNN network to segment nanoparticle images and optimizing the network to improve accuracy, the size and shape parameters of the nanoparticles were successfully measured.
SIGNAL IMAGE AND VIDEO PROCESSING
(2021)
Article
Computer Science, Artificial Intelligence
Yanbei Liu, Lianxi Fan, Changqing Zhang, Tao Zhou, Zhitao Xiao, Lei Geng, Dinggang Shen
Summary: Alzheimers disease is a complex neurodegenerative disease, and early diagnosis and treatment have been major concerns. Researchers have proposed a new Auto-Encoder based Multi-View missing data Completion framework for learning common representations for AD diagnosis effectively.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Materials Science, Textiles
Jun Wu, Da Wang, Zhitao Xiao, Kun Yu, Fang Zhang, Lei Geng
Summary: The study proposed an objective pilling rating method based on CNN, established a pilling image dataset of various fabrics and hairball shapes, and achieved a rating accuracy of 97.70% with the SONet rating system model.
JOURNAL OF THE TEXTILE INSTITUTE
(2022)
Article
Engineering, Electrical & Electronic
Zhitao Xiao, Kai Yin, Lei Geng, Jun Wu, Fang Zhang, Yanbei Liu
Summary: This paper proposes an end-to-end pest identification network that combines deep learning and hyperspectral imaging technology for effective pest control. The method utilizes spectral feature extraction and spectral-spatial feature extraction to enhance pest identification accuracy and proves to be more suitable for pest identification tasks than other methods.
SIGNAL IMAGE AND VIDEO PROCESSING
(2022)
Article
Engineering, Electrical & Electronic
Lei Geng, Quan Guo, Zhitao Xiao, Jun Tong, Yuelong Li
Summary: Accurately classifying dead embryos and live embryos is crucial for the successful development of vaccines. In this study, we utilized deep learning techniques and photoplethysmographic waveform to detect embryo activity. By rescaling the data and constructing a novel detection model, we achieved equal treatment of each feature in the data and powerful feature extraction capabilities.
SIGNAL IMAGE AND VIDEO PROCESSING
(2022)
Article
Materials Science, Textiles
Yongmin Guo, Zhitao Xiao, Lei Geng
Summary: This paper proposes a method based on SE-PSPNet for defect detection in 3D braided composites. The method addresses the poor segmentation effect of small objects with inclusion defect by introducing the SE module and setting different weights. Experimental results show that the proposed method outperforms other models on different evaluation metrics.
JOURNAL OF THE TEXTILE INSTITUTE
(2023)
Article
Engineering, Electrical & Electronic
Jun Wu, Yaxin Zhang, Zhitao Xiao, Fang Zhang, Lei Geng
Summary: In this paper, an attention mechanism based on residual convolution module U-Net (RCU-Net) is proposed for the automatic segmentation of the retinal layer and cystoid edema lesions in DME. By fusing the residual structure and CBAM for feature extraction, the network can effectively learn different levels of information. Experimental results show that the proposed method achieves high accuracy and MIoU in DME segmentation.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Peripheral Vascular Disease
Lei Chen, Shuncong Wang, Yuanbo Feng, Jie Yu, Walter Coudyzer, Chantal Van Ongeval, Lei Geng, Yue Li, Yicheng Ni
Summary: This study established an inovo platform combining laser speckle contrast imaging (LSCI) and chicken embryo chorioallantoic membrane (CAM) to monitor vascular diameters and perfusion in real-time and investigate the off-targeting effects of tumor vascular disrupting agents (VDAs) on normal vessels. The results showed that ED12 was the optimal time window for studying vasoactive drugs. This platform provides a powerful tool for further research on VDAs and other vasoactive medications.
MICROVASCULAR RESEARCH
(2022)
Article
Mathematics, Applied
Lei Geng, Muhammad Shoaib Saleem, Kiran Naseem Aslam, Rahat Bano
Summary: This paper deals with fractional integral inequalities for strongly reciprocally p,h-convex functions and establishes related inequalities, extending existing research results.
JOURNAL OF FUNCTION SPACES
(2022)
Article
Computer Science, Information Systems
Yanbei Liu, Shichuan Zhao, Xiao Wang, Lei Geng, Zhitao Xiao, Jerry Chun-Wei Lin
Summary: Graph Neural Networks (GNNs), based on deep learning, have attracted research interest. Many GNNs have achieved state-of-the-art accuracy but lack supervision information for unlabeled data. To address this, we propose SCGNN which extracts self-supervision information from unlabeled nodes and utilizes label information from labeled nodes. Experimental results show that SCGNN outperforms baselines, improving accuracy by an average of 2.08% and by 5.8% on the Disease dataset.
IEEE TRANSACTIONS ON BIG DATA
(2023)
Article
Computer Science, Artificial Intelligence
Yanbei Liu, Wanjin Shan, Xiao Wang, Zhitao Xiao, Lei Geng, Fang Zhang, Dongdong Du, Yanwei Pang
Summary: This paper proposes a novel framework called Cross-scale Contrastive Triplet Networks (CCTN) for graph representation learning. By capturing both contextual and intrinsic node information through contrastive learning, CCTN achieves state-of-the-art performance on node classification and clustering tasks.
PATTERN RECOGNITION
(2024)
Proceedings Paper
Computer Science, Artificial Intelligence
Jun Wu, Jianxu Chen, Zhitao Xiao, Lei Geng
Summary: The paper proposes an end-to-end automatic retinal layer segmentation method called DA-PSPNet, based on deep learning, which can accurately segment seven retinal layers in OCT images. The method integrates a dual attention mechanism to extract richer layer boundary information and achieves better performance compared to traditional methods.
PROCEEDINGS OF 2021 3RD INTERNATIONAL CONFERENCE ON INTELLIGENT MEDICINE AND IMAGE PROCESSING (IMIP 2021)
(2021)