Article
Computer Science, Artificial Intelligence
Amirhossein Aghamohammadi, Ramin Ranjbarzadeh, Fatemeh Naiemi, Marzieh Mogharrebi, Shadi Dorosti, Malika Bendechache
Summary: The paper presents a strategy based on cascade convolutional neural network to address the challenges of liver and tumor segmentation. The Z-Score algorithm is used for image normalization and a novel encoding algorithm LDOG is proposed for key feature recognition. A cascade CNN structure is employed to extract local and semi-global features, utilizing a simple but effective model to improve segmentation accuracy and efficiency, outperforming current state-of-the-art works.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Vanda Czipczer, Andrea Manno-Kovacs
Summary: Liver plays a crucial role in metabolic processes, making fast diagnosis and surgical planning vital. This study introduces an automatic, deep learning based approach for liver segmentation that is adaptable and can handle smaller databases, resulting in improved segmentation accuracy.
Article
Computer Science, Artificial Intelligence
Ramin Ranjbarzadeh, Shadi Dorosti, Saeid Jafarzadeh Ghoushchi, Sadaf Safavi, Navid Razmjooy, Nazanin Tataei Sarshar, Shokofeh Anari, Malika Bendechache
Summary: This research aims to diagnose ICP quickly using a computer-aided diagnosis system. By employing methods such as image segmentation, image quality enhancement, and key information representation, the accurate detection of the nerve optic regions and calculation of its diameter were successfully achieved. Experimental results showed that compared to other methods, this approach can more accurately detect ICP, which can be used for early diagnosis and prevention of health problems in patients.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Chemistry, Analytical
Kh Tohidul Islam, Sudanthi Wijewickrema, Stephen O'Leary
Summary: This paper proposes a solution for brain tumor segmentation, involving enhancing MRI dataset with synthetic CT images and optimizing CNN architecture. Experimental results demonstrate the superiority of the proposed method over existing methods.
Article
Computer Science, Artificial Intelligence
Pradip Paithane, Sangeeta Kakarwal
Summary: The LMNS-net deep learning model is a fast and accurate approach for automatic pancreatic segmentation in clinical abdominal CT images. It utilizes a lightweight multiscale module to reduce computation time and achieve high accuracy. The model takes only 1-3 seconds for segmentation in the testing process, making it faster and more efficient than other approaches.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Nanyan Shen, Ziyan Wang, Jing Li, Huayu Gao, Wei Lu, Peng Hu, Lanyun Feng
Summary: A segmentation model based on U-Net is proposed for multi-organ segmentation in hepato-biliary-pancreatic surgery, with deformable receptive fields and a spatial attention block. A deformable convolution block is used to handle variations in shapes and sizes. The skip-connection structure of U-Net is improved with multi-scale attention maps and high-level semantic information. Experimental results show that the proposed model achieves better segmentation performance compared to U-Net on the TCIA multiorgan segmentation dataset.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Mathematics
Giovanna Maria Dimitri, Paolo Andreini, Simone Bonechi, Monica Bianchini, Alessandro Mecocci, Franco Scarselli, Alberto Zacchi, Guido Garosi, Thomas Marcuzzo, Sergio Antonio Tripodi
Summary: Deep learning techniques are applied in bioinformatics and biomedical imaging for the automatic identification and segmentation of glomeruli in kidney tissues. The results show promising performance in the segmentation task and the proposed use of the CD10 staining procedure for sclerotic glomeruli segmentation is effective.
Article
Biochemical Research Methods
Jason Causey, Jonathan Stubblefield, Jake Qualls, Jennifer Fowler, Lingrui Cai, Karl Walker, Yuanfang Guan, Xiuzhen Huang
Summary: This paper presents the solution and results of the Arkansas AI-Campus team in the 2019 Kidney Tumor Segmentation Challenge. Their deep learning model achieved high Dice scores in kidney and tumor segmentation and secured a good ranking in the competition.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Computer Science, Information Systems
Simone Bonechi, Paolo Andreini, Alessandro Mecocci, Nicola Giannelli, Franco Scarselli, Eugenio Neri, Monica Bianchini, Giovanna Maria Dimitri
Summary: The automatic segmentation of the aorta using 2D convolutional neural networks and 3D CT scans as input is presented in this paper. A semi-automated approach was used to obtain 3D annotations for a set of CT images, and two different network architectures were compared for segmentation on three CT views. The results show promising accuracy and efficiency of the neural networks in providing aortic segmentation.
Article
Biology
Saeed Mohagheghi, Amir Hossein Foruzan
Summary: The study introduced an explainable deep correction method that refined the output of other models with cascaded 1D and 2D models to provide reliable and accurate results. The proposed method achieved state-of-the-art results in 3D liver segmentation tasks.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Biology
Guoting Luo, Qing Yang, Tao Chen, Tao Zheng, Wei Xie, Huaiqiang Sun
Summary: In this study, a novel two-stage deep neural network for adrenal gland segmentation was proposed. The cascaded framework outperforms the state-of-the-art method in terms of accuracy, requires fewer trainable parameters, and imposes a smaller demand on computational resources.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Engineering, Biomedical
Li Kang, Ziqi Zhou, Jianjun Huang, Wenzhong Han
Summary: A deep learning-based segmentation method for renal tumors is proposed in this study, utilizing prior information and ConvLSTM to improve segmentation accuracy. Experimental results demonstrate superior performance on the Kits19 dataset.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Article
Oncology
Xiaobo Wen, Bing Liang, Biao Zhao, Xiaokun Hu, Meifang Yuan, Wenchao Hu, Ting Liu, Yi Yang, Dongming Xing
Summary: The objective of this study was to improve the automated segmentation of temporal lobes on CT images for radiotherapy by finding a new loss function and addressing the issue of class-imbalanced samples. The results showed that the U-Net model based on FGD-BCEL outperformed the other four loss function-based models in terms of evaluation metrics, and it produced segmentations that were morphologically closer to the ground truth mask maps.
FRONTIERS IN ONCOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Jianning Chi, Shuang Zhang, Xiaoying Han, Huan Wang, Chengdong Wu, Xiaosheng Yu
Summary: The study proposes a multi-input directional UNet (MID-UNet) for segmenting COVID-19 infections in lung CT images, addressing the problems faced by current deep learning methods in this field. Experimental results demonstrate that the proposed method exhibits superior performance in segmenting COVID-19 infections.
SIGNAL PROCESSING-IMAGE COMMUNICATION
(2022)
Article
Computer Science, Interdisciplinary Applications
Yu Deng, Ling Wang, Chen Zhao, Shaojie Tang, Xiaoguang Cheng, Hong-Wen Deng, Weihua Zhou
Summary: This study proposed a deep learning-based approach for the fast and automatic extraction of the periosteal and endosteal contours of the proximal femur. The developed convolutional neural network (CNN) achieved high accuracy in segmentation, with similarity coefficients exceeding 96%. This method significantly reduces segmentation time and shows potential in assessing bone density and strength.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2022)
Article
Chemistry, Multidisciplinary
Xue Wang, Li Zhang, Meiling Xiao, Junjie Ge, Wei Xing, Changpeng Liu, Jianbing Zhu
Summary: A polymerchelation strategy is used to disperse Fe-Nx active sites onto the carbon surface, resulting in a hierarchically porous structure and excellent conductivity. The optimal catalyst exhibits impressive oxygen reduction reaction activity, surpassing the Pt/C benchmark.
CHINESE CHEMICAL LETTERS
(2023)
Article
Cell Biology
Qian Wang, Zhaoyang Chen, Junjie Guo, Xiaoping Peng, Zeqi Zheng, Hang Chen, Haibo Liu, Yuanji Ma, Jianbing Zhu
Summary: Our study found that atorvastatin-induced tDCs have a beneficial effect on post-infarction cardiomyocyte apoptosis and myocardial fibrosis, reducing inflammatory cell infiltration and inhibiting oxidative stress, likely through suppression of TLR-4/NF-kappa B activation.
INFLAMMATION RESEARCH
(2023)
Review
Chemistry, Multidisciplinary
Jinsheng Li, Changpeng Liu, Junjie Ge, Wei Xing, Jianbing Zhu
Summary: Alkaline hydrogen-electricity energy conversion technologies, such as AEMFCs and AEMWEs, offer advantages over acidic counterparts, but face challenges in terms of AEM properties. This review examines the main obstacles of AEMs, including ion conductivity, stability, and device integration, and proposes strategies to address these challenges. The insights provided by this review can accelerate the commercialization of these promising hydrogen-electric energy conversion technologies.
CHEMISTRY-A EUROPEAN JOURNAL
(2023)
Article
Geriatrics & Gerontology
Shouyun Du, Yiqing Wang, Guodong Li, Hongyu Wei, Hongjie Yan, Xiaojing Li, Yijie Wu, Jianbing Zhu, Yi Wang, Zenglin Cai, Nizhuan Wang
Summary: This study analyzed the resting-state functional magnetic resonance data of a Chinese population comprising 14 patients with PD and 13 controls using the functional covariance connection strength method. It was found that patients with PD had abnormal connections between olfactory-related brain regions and white matter fiber bundles, which were associated with the symptoms of Parkinson's disease.
FRONTIERS IN AGING NEUROSCIENCE
(2023)
Article
Nanoscience & Nanotechnology
Di Yang, Jinsheng Li, Changpeng Liu, Junjie Ge, Wei Xing, Jianbing Zhu
Summary: In this study, a MXene/Zn metal anode interfacial structure with a protective layer of single/few-layer Ti3C2Tx MXene is designed to address the challenges faced by Zn metal anodes in alkaline electrolytes. The MXene layer isolates the direct contact between the Zn metal anodes and the electrolytes and inhibits zincate dissolution, resulting in improved cycle stability. The Ti3C2Tx-protected Zn metal anode demonstrates superior performance compared to the bare Zn counterpart, with stable operation for over 400 cycles at a high current density of 5.0 mA cm-2.
ACS APPLIED MATERIALS & INTERFACES
(2023)
Article
Cell Biology
Wei Dong, Chen Dong, Jianbing Zhu, Yaofu Zheng, Junfei Weng, Leilei Liu, Yang Ruan, Xu Fang, Jin Chen, Wenyu Liu, Xiaoping Peng, Xuanying Chen
Summary: Myocardial ischemia/reperfusion injury (MIRI) is a major cause of heart failure after myocardial infarction. MiR-322 plays a role in regulating MIRI progression, and its mechanism involves the interaction between Smad7/Smurf2, HIF-1 alpha, and beta-catenin. The study demonstrates that upregulating miR-322 improves MIRI by activating the Smad3/beta-catenin pathway through targeting Smurf2 and Smad7, while the positive feedback loop between beta-catenin and HIF-1 alpha continuously enhances MIRI.
CELL BIOLOGY INTERNATIONAL
(2023)
Article
Engineering, Electrical & Electronic
Wenyuan Zhong, Huaxiong Li, Qinghua Hu, Yang Gao, Chunlin Chen
Summary: Deep learning methods have attracted much attention for image classification recently. However, for small-scale data, these methods may not yield optimal results due to the lack of training samples. Sparse representation is efficient and interpretable, but its precision is not competitive. To address this issue, we propose a Multi-Level Cascade Sparse Representation (ML-CSR) learning method that combines the advantages of both deep learning and sparse representation. ML-CSR utilizes a pyramid structure and two core modules, Error-To-Feature (ETF) and Generate-Adaptive-Weight (GAW), to improve precision. Experiments on face databases demonstrate the effectiveness of ML-CSR, and ablation experiments further confirm the benefits of the proposed pyramid structure, ETF, and GAW modules. The code is available at https://github.com/Zhongwenyuan98/ML-CSR.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Computer Science, Interdisciplinary Applications
Shumeng Li, Heng Cai, Lei Qi, Qian Yu, Yinghuan Shi, Yang Gao
Summary: This paper proposes a novel barely-supervised segmentation setting with few sparsely-labeled images and a large amount of unlabeled images. By introducing a parasitic-like network, the collaboration of two modules is achieved through three stages of infection, development, and eclosion, providing accurate pseudo-labels for training. The results demonstrate that the framework achieves high performance on extremely sparse annotation tasks.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Multidisciplinary Sciences
Zhaoping Shi, Ji Li, Yibo Wang, Shiwei Liu, Jianbing Zhu, Jiahao Yang, Xian Wang, Jing Ni, Zheng Jiang, Lijuan Zhang, Ying Wang, Changpeng Liu, Wei Xing, Junjie Ge
Summary: The reaction route plays a crucial role in determining the stability and catalytic performance of ruthenium-based catalysts. By controlling the charge of ruthenium, the reaction route can be customized, leading to improved stability and lifespan of the catalysts in electrolyzers.
NATURE COMMUNICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Xiao Liu, Shuyang Liu, Bo An, Yang Gao, Shangdong Yang, Wenbin Li
Summary: Interpretable policy distillation aims to convert a deep reinforcement learning policy into a self-explainable model, but it often fails to perform well on complex tasks. This research identifies the heavy-tailed nature of the experience distributions as a critical issue and proposes a method to characterize decision boundaries using minimum experience retention, resulting in improved distilled policies.
IEEE INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yanfang Liu, Xiaocong Fan, Wenbin Li, Yang Gao
Summary: This study combines active query strategy and passive-aggressive update strategy, and proposes a novel online active learning algorithm for trapezoidal data streams. Experimental results confirm its effectiveness in learning from trapezoidal data streams.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Review
Chemistry, Applied
Di Yang, Jinsheng Li, Changpeng Liu, Wei Xing, Jianbing Zhu
Summary: Alkaline Zn-based primary batteries have been successful, but secondary batteries using Zn anodes face challenges due to poor cycle reversibility. Various degradation mechanisms contribute to the failures, and their interactions make it difficult to address all the issues with a single strategy. Therefore, a comprehensive evaluation of different strategies is important to commercialize alkaline Zn batteries. This review systematically analyzes the progress and performance of improvement strategies for Zn anodes in alkaline conditions, highlighting design strategies from the perspectives of ion and electron regulation and comparing their advantages and disadvantages based on comprehensive performance parameters.
JOURNAL OF ENERGY CHEMISTRY
(2023)
Article
Engineering, Electrical & Electronic
Wenjun Lv, Chenhui Yuan, Jichen Wang, Jianbing Zhu, Yu Kang, Ji Chang
Summary: Geophysical logging instruments provide a feasible way to model fine borehole geology by continuously measuring multiple geophysical properties of borehole rocks. Machine learning has been demonstrated effective in generating missing well logs, but the independent and identical distribution assumption is not satisfied in the case of cross-well missing logs generation. To address this issue, we propose an explainable regression network named LogRegX, which integrates feature extraction, alignment, and missing logs prediction while maintaining feature explainability.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Proceedings Paper
Operations Research & Management Science
Xin Liu, Lu Han, Jian-Bing Zhu, Yong-Quan Wei
Summary: This paper introduces a new method to solve the problem of traditional spacecraft health state monitoring, which utilizes an on-board health monitoring embedded software to judge the spacecraft's health state based on real-time telemetry data and monitoring rules, and automatically handles anomalies. The software has been successfully applied in China Space Station and a remote sensing spacecraft agile platform, significantly improving the reliability of the spacecraft and reducing the cost of ground health monitoring.
APPLICATIONS OF DECISION SCIENCE IN MANAGEMENT, ICDSM 2022
(2023)