检索结果-内蒙古大学图书馆

Deep and handcrafted feature supported diabetic retinopathy detection: A study

学校读者我要写书评

暂无评论

Procedia Computer Science 2023年 218卷 2675-2683页

作者： Seifedine Kadry Rubén González Crespo Enrique Herrera-Viedma Sujatha Krishnamoorthy Venkatesan Rajinikanth Faculty of Applied Computing and Technology Noroff University College Kristiansand 94612 Norway Computer Science Department School of Engineering and Technology Universidad Internacional de La Rioja 26006 Andalusian Spain Research Institute in Data Science and Computational Intelligence University of Granada Granada Spain Zhejiang Bioinformatics International Science and Technology Cooperation Center Wenzhou-Kean University Zhejiang Province China Wenzhou Municipal Key Lab of Applied Biomedical and Biopharmaceutical Informatics Wenzhou-Kean University Zhejiang Province China Department of Electronics and Instrumentation Engineering St. Joseph's College of Engineering OMR Chennai 600119 Tamil Nadu India

The eye is the prime sensory organ in physiology, and the abnormality in the eye severely influences the vision system. Therefore, eye irregularity is commonly assessed using imaging schemes, and Fundus Retinal Image (FRI) supported eye screening is one of the ophthalmological practices. This work proposed a Deep-Learning Procedure (DLP) to recognize Diabetic Retinopathy (DR) in FI. The proposed work presents the experimental work with different DLP methods found in the literature. This work is executed with two modes; (i) DR detection using conventional deep-features and (ii) DR discovery using deep ensemble features. To demonstrate this work, 1800 fundus images (900 regular and 900 DR class) are considered for the assessment, and the advantage of proposed plan is confirmed using various performance metrics. The experimental outcome of this study confirms that the AlexNet-based detection provides a better detection (>96%), and the deep ensemble features of AlexNet, VGG16, and ResNet18 provide a detection accuracy of >98% on the chosen FRI database.

关键词： Eye abnormality Diabetic retinopathy Fundus imaging Deep-Learning Classification

Automatic detection of lung nodule in CT scan slices using CNN segmentation schemes: A study

学校读者我要写书评

暂无评论

Procedia Computer Science 2023年 218卷 2786-2794页

作者： Seifedine Kadry Enrique Herrera-Viedma Rubén González Crespo Sujatha Krishnamoorthy Venkatesan Rajinikanth Faculty of Applied Computing and Technology Noroff University College Kristiansand 94612 Norway Research Institute in Data Science and Computational Intelligence University of Granada Granada Andalusian Spain Computer Science Department School of Engineering and Technology Universidad Internacional de La Rioja 26006 Spain Zhejiang Bioinformatics International Science and Technology Cooperation Center Wenzhou-Kean University Zhejiang Province China Wenzhou Municipal Key Lab of Applied Biomedical and Biopharmaceutical Informatics Wenzhou-Kean University Zhejiang Province China Department of Electronics and Instrumentation Engineering St. Joseph's College of Engineering OMR Chennai 600119 Tamil Nadu India

The lung is one of the prime respiratory organs in human physiology, and its abnormality will severely disrupt the respiratory system. Lung Nodule (LN) is one of the abnormalities, and early screening and treatment are necessary to reduce its harshness. The proposed work aims to implement the Convolutional-Neural-Network (CNN) segmentation methodology to extract the LN in various lung CT slices, such as axial, coronal, and sagittal planes. This work consists of the following phases; (i) Image collection and pre-processing, (ii) Ground-truth generation, (iii) CNN-supported segmentation, and (iv) Performance evaluation and validation. In this work, the merit of pre-trained CNN segmentation schemes is verified using (i) One-fold training and (ii) Two-fold training methods. The test images for this study are collected from The Cancer Imaging Archive (TCIA) database. The experimental investigation is executed using Python®, and the outcome of this study confirms that the VGG-SegNet helps to get better values of Jaccard (>88%), Dice (>93%), and Accuracy (>96%) compared to other CNN methods.

关键词： Lung nodule CT scan CNN segmentation VGG-SegNet Accuracy Detection

USL-Net: Uncertainty Self-Learning Network for Unsupervised Skin Lesion Segmentation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Xiaofan Peng, Bo Hu, Jie Ma, Changyou Yang, Daipeng Xie, Zhuyang School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu611756 China Engineering Research Center of Sustainable Urban Intelligent Transportation Ministry of Education Chengdu611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong University Chengdu611756 China Manufacturing Industry Chains Collaboration and Information Support Technology Key Laboratory of Sichuan Province Southwest Jiaotong University Chengdu611756 China Data Recovery Key Laboratory of Sichuan Province Neijiang Normal University Sichuan Neijiang641100 China

Unsupervised skin lesion segmentation offers several benefits, such as conserving expert human resources, reducing discrepancies caused by subjective human labeling, and adapting to novel environments. However, segmenting dermoscopic images without manual labeling guidance is a challenging task due to image artifacts such as hair noise, blister noise, and subtle edge differences. In this paper, we introduce an innovative Uncertainty Self-Learning Network (USL-Net) to eliminate the need for manual labeling guidance for the segmentation. Initially, features are extracted using contrastive learning, followed by the generation of Class Activation Maps (CAMs) as saliency maps. High-saliency regions in the map serve as pseudo-labels for lesion regions while low-saliency regions represent the background. Besides, intermediate regions can be hard to classify, often due to their proximity to lesion edges or interference from hair or blisters. Rather than risking potential pseudo-labeling errors or learning confusion by forcefully classifying these regions, they are taken as uncertainty regions by exempted from pseudo-labeling and allowing the network to self-learning. Further, we employ connectivity detection and centrality detection to refine foreground pseudo-labels and reduce noise-induced errors. The performance is further enhanced by the iterated refinement process. The experimental validation on ISIC-2017, ISIC-2018, and PH2 datasets demonstrates that its performance is comparable to supervised methods, and exceeds that of other existing unsupervised methods. On the typical ISIC-2017 dataset, our method outperforms state-of-the-art unsupervised methods by 1.7% in accuracy, 6.6% in Dice coefficient, 4.0% in Jaccard index, and 10.6% in sensitivity. Copyright © 2023, The Authors. All rights reserved.

关键词： Contrastive Learning

ScalaBFS: A scalable BFS accelerator on HBM-enhanced FPGAs

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Li, Kexin Liu, Chenhao Shao, Zhiyuan Wang, Zeke Wu, Minkang Chen, Jiajie Liao, Xiaofei Jin, Hai National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan30074 China Collaborative Innovation Center of Artificial Intelligence Zhejiang University China

High Bandwidth Memory (HBM) provides massive aggregated memory bandwidth by exposing multiple memory channels to the processing units. To achieve high performance, an accelerator built on top of an FPGA configured with HBM (i.e., FPGA-HBM platform) needs to scale its performance according to the available memory channels. In this paper, we propose an accelerator for BFS (Breadth-First Search) algorithm, named as ScalaBFS, that builds multiple processing elements to sufficiently exploit the high bandwidth of HBM to improve efficiency. We implement the prototype system of ScalaBFS and conduct BFS in both real-world and synthetic scale-free graphs on Xilinx Alveo U280 FPGA card real hardware. The experimental results show that ScalaBFS scales its performance almost linearly according to the available memory pseudo channels (PCs) from the HBM2 subsystem of U280. By fully using the 32 PCs and building 64 processing elements (PEs) on U280, ScalaBFS achieves a performance up to 19.7 GTEPS (Giga Traversed Edges Per Second). When conducting BFS in sparse real-world graphs, ScalaBFS achieves equivalent GTEPS to Gunrock running on the state-of-art Nvidia V100 GPU that features 64-PC HBM2 (twice memory bandwidth than U280). Copyright © 2021, The Authors. All rights reserved.

关键词： Field programmable gate arrays (FPGA)

Neuromodulated Meta-Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Jingyao Guo, Huijie Qiang, Wenwen Li, Jiangmeng Zheng, Changwen Xiong, Hui Hua, Gang University of Chinese Academy of Sciences Beijing China National Key Laboratory of Space Integrated Information System Institute of Software Chinese Academy of Sciences Beijing China Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing China Thrust of Artificial Intelligence Hong Kong University of Science and Technology Guangzhou China Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong Wormpex AI Research LLC BellevueWA98004 United States

Humans excel at adapting perceptions and actions to diverse environments, enabling efficient interaction with the external world. This adaptive capability relies on the biological nervous system (BNS), which activates different brain regions for distinct tasks. Meta-learning similarly trains machines to handle multiple tasks but relies on a fixed network structure, not as flexible as BNS. To investigate the role of flexible network structure (FNS) in meta-learning, we conduct extensive empirical and theoretical analyses, finding that model performance is tied to structure, with no universally optimal pattern across tasks. This reveals the crucial role of FNS in meta-learning, ensuring meta-learning to generate the optimal structure for each task, thereby maximizing the performance and learning efficiency of meta-learning. Motivated by this insight, we propose to define, measure, and model FNS in meta-learning. First, we define that an effective FNS should possess frugality, plasticity, and sensitivity. Then, to quantify FNS in practice, we present three measurements for these properties, collectively forming the structure constraint with theoretical supports. Building on this, we finally propose Neuromodulated Meta-Learning (NeuronML) to model FNS in meta-learning. It utilizes bi-level optimization to update both weights and structure with the structure constraint. Extensive theoretical and empirical evaluations demonstrate the effectiveness of NeuronML on various tasks. Code is publicly available at https://***/WangJingyao07/NeuronML. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

Unlocking the Secrets Behind Advanced Artificial intelligence Language Models in Deidentifying Chinese-English Mixed Clinical Text: Development and Validation Study

学校读者我要写书评

暂无评论

Journal of Medical Internet research 2024年第1期26卷 e48443页

作者： Lee, You-Qian Chen, Ching-Tai Chen, Chien-Chang Lee, Chung-Hong Chen, Peitsz Wu, Chi-Shin Dai, Hong-Jie Dialogue System Technical Department Asustek Computer Inc Taipei Taiwan Intelligent System Laboratory Department of Electrical Engineering College of Electrical Engineering and Computer Science National Kaohsiung University of Science and Technology Kaohsiung Taiwan Department of Bioinformatics and Medical Engineering Asia University Taichung Taiwan Center for Precision Health Research Asia University Taichung Taiwan Electromagnetic Sensing Control and AI Computing System Laboratory Department of Electrical Engineering College of Electrical Engineering and Computer Science National Kaohsiung University of Science and Technology Kaohsiung Taiwan Knowledge Discovery and Data Mining Lab Department of Electrical Engineering College of Electrical Engineering and Computer Science National Kaohsiung University of Science and Technology Kaohsiung Taiwan Department of Chemical Engineering Feng Chia University Taichung Taiwan National Center for Geriatrics and Welfare Research National Health Research Institutes Zhunan Taiwan National Institute of Cancer Research National Health Research Institutes Tainan Taiwan School of Post-Baccalaureate Medicine College of Medicine Kaohsiung Medical University Kaohsiung Taiwan Center for Big Data Research Kaohsiung Medical University Kaohsiung Taiwan

Background: The widespread use of electronic health records in the clinical and biomedical fields makes the removal of protected health information (PHI) essential to maintain privacy. However, a significant portion of information is recorded in unstructured textual forms, posing a challenge for deidentification. In multilingual countries, medical records could be written in a mixture of more than one language, referred to as code mixing. Most current clinical natural language processing techniques are designed for monolingual text, and there is a need to address the deidentification of code-mixed text. Objective: The aim of this study was to investigate the effectiveness and underlying mechanism of fine-tuned pretrained language models (PLMs) in identifying PHI in the code-mixed context. Additionally, we aimed to evaluate the potential of prompting large language models (LLMs) for recognizing PHI in a zero-shot manner. Methods: We compiled the first clinical code-mixed deidentification data set consisting of text written in Chinese and English. We explored the effectiveness of fine-tuned PLMs for recognizing PHI in code-mixed content, with a focus on whether PLMs exploit naming regularity and mention coverage to achieve superior performance, by probing the developed models’ outputs to examine their decision-making process. Furthermore, we investigated the potential of prompt-based in-context learning of LLMs for recognizing PHI in code-mixed text. Results: The developed methods were evaluated on a code-mixed deidentification corpus of 1700 discharge summaries. We observed that different PHI types had preferences in their occurrences within the different types of language-mixed sentences, and PLMs could effectively recognize PHI by exploiting the learned name regularity. However, the models may exhibit suboptimal results when regularity is weak or mentions contain unknown words that the representations cannot generate well. We also found that the availability of cod

关键词： ChatGPT code mixing deidentification electronic health record large language model pretrained language model

Locate who you are: Matching geo-location to text for user identity linkage

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Shao, Jiangli Wang, Yongqing Gao, Hao Shen, Huawei Li, Yangyang Cheng, Xueqi Data Intelligence System Research Center Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences National Engineering Lab for Risk Perception and Prevention China Academic Electronic and Information Technology

Nowadays, users are encouraged to activate across multiple online social networks simultaneously. User identity linkage, which aims to reveal the correspondence among different accounts across networks, has been regarded as a fundamental problem for user profiling, marketing, cybersecurity, and recommendation. Existing methods mainly address the prediction problem by utilizing profile, content, or structural features of users in symmetric ways. However, encouraged by online services, information from different social platforms may also be asymmetric, such as geo-locations and texts. It leads to an emerged challenge in aligning users with asymmetric information across networks. Instead of similarity evaluation applied in previous works, we formalize correlation between geo-locations and texts and propose a novel user identity linkage framework for matching users across networks. Moreover, our model can alleviate the label scarcity problem by introducing external text-location pairs. Experimental results on real-world datasets show that our approach outperforms existing methods and achieves state-of-the-art results. © 2021, CC BY-NC-ND.

关键词： Social networking (online)

Reconstructing Missing Modalities in Multi-Modal Endoscopic Ultrasound Via Cross-Modal Feature Replacement Representation

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Zheng, Cenyang Gong, Xun Fan, Lin Li, Jiao School of Computing and Artificial Intelligence Southwest Jiaotong University Sichuan Chengdu611756 China Engineering Research Center of Sustainable Urban Intelligent Transportation Ministry of Education Chengdu611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong University Chengdu611756 China Manufacturing Industry Chains Collaboration and Information Support Technology Key Laboratory of Sichuan Province Southwest Jiaotong University Chengdu611756 China Department of Gastroenterology The Third People’s Hospital of Chendu Affiliated Hospital of Southwest Jiaotong University Chengdu610031 China

Endoscopic ultrasound (EUS) is the primary diagnostic imaging technique for gastrointestinal stromal tumors. However, accurately identifying these tumors through EUS imaging alone remains challenging, even for experienced clinicians. They often rely on additional data, such as white light endoscopy (WLE) and EUS text reports, to aid in the diagnosis. Unfortunately, in practice, these supplementary modalities are sometimes overlooked during patient evaluation. To address these issues, we propose the Cross-modal Feature Reconstruction Network (CFRN), which integrates multimodal data modeling for medical diagnostic processes and performs specific learning for different modalities with missing data. The network fuses the multi-modal features based on the clinical medical process, realizes the feature-level missing modality reconstruction, and learns the refined features of different modal true cases by decoding the mode-specific missing. We trained and tested the model on two publicly available datasets, Audiovision-MNIST, MM-IMDb, and data sets collected from real clinical scenarios at three hospitals. Our results show that CFRN outperforms state-of-the-art methods, achieving higher prediction accuracy than using EUS alone. By simultaneously incorporating multimodal or reconstructed modal data for disease diagnosis, our single-modal EUS model proves to be more suitable for clinical scenarios. © 2024, The Authors. All rights reserved.

关键词： Endoscopy

Shape-aware contrastive deep supervision for esophageal tumor segmentation from CT scans

学校读者我要写书评

暂无评论

Shape-aware contrastive deep supervision for esophageal tumo...

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Jin, Qiangguo Cui, Hui Sun, Changming Huang, Jiapeng Xuan, Ping Xu, Yiyue Wang, Linlin Cao, Leilei Wei, Leyi Su, Ran Northwestern Polytechnical University School of Software Shaanxi China Yangtze River Delta Research Institute of Northwestern Polytechnical University Taicang China La Trobe University Department of Computer Science and Information Technology Melbourne Australia Csiro Data61 Sydney Australia Shantou University School of Engineering Department of Computer Science Guangdong China Shandong First Medical University Shandong Academy of Medical Sciences Shandong Cancer Hospital and Institute Department of Radiation Oncology Shandong China Zhejiang University Innovation Center of Yangtze River Delta Zhejiang China Shandong University School of Software Shandong China Tianjin University School of Computer Software College of Intelligence and Computing Tianjin China

ISBN: (纸本)9798350337488

Accurate tumor segmentation is crucial for esophageal cancer radiotherapy treatment planning. The low contrast among the esophagus, tumors, and surrounding tissues, and irregular tumor shapes limit the performance of automatic segmentation methods. In this paper, we aim to exploit the irregular shapes of tumors to facilitate accurate segmentation. We propose a simple and pluggable shape-aware contrastive deep supervision network (SCDSNet) with shape-aware regularization and voxel-to-voxel contrastive deep supervision. Specifically, the shape-aware regularization with an uncertainty minimization strategy encourages the precise predictions of an additional shape-aware head. The voxel-to-voxel contrastive deep supervision enhances the multi-scale shape-tumor contrast for better voxel-to-voxel prediction of shapes. The proposed method is simple and highly pluggable, which can easily be extended to other frameworks. Further, we establish a large in-house dataset on esophageal cancer to validate the effectiveness of our proposed method. The quantitative and qualitative experimental results demonstrate the effectiveness of SCDSNet on the esophageal cancer dataset. © 2023 IEEE.

关键词： Deep supervision Esophagus tumor segmentation Shape-aware regularization Voxel-to-voxel contrastive learning