Article
Computer Science, Hardware & Architecture
Mahmoud Ragab, Wajdi H. Aljedaibi, Alaa F. Nahhas, Ibrahim R. Alzahrani
Summary: This paper presents a deep learning-based computer aided diagnosis model for diabetic retinopathy detection and grading. The model includes preprocessing, image segmentation, feature extraction, and classification, and achieves high accuracy according to the performance validation on benchmark data set.
COMPUTERS & ELECTRICAL ENGINEERING
(2022)
Review
Engineering, Biomedical
K. C. Pavithra, Preetham Kumar, M. Geetha, Sulatha V. Bhandary
Summary: Diabetic Macular Edema (DME) is a serious complication of Diabetic Retinopathy (DR) and it is the leading cause of vision loss in diabetics. DME is characterized by the accumulation of fluid in the macula due to leaky blood vessels. Advanced imaging techniques such as Color Fundus Photography (CFP) and Optical Coherence Tomography (OCT) can detect the presence of DME at different stages of DR. This review article discusses the latest automated DME detection methods using traditional Machine Learning (ML) and Deep Learning (DL) techniques with retinal fundus or OCT images.
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING
(2023)
Article
Engineering, Biomedical
Rui Liu, Qingchen Li, Feiping Xu, Shasha Wang, Jie He, Yiting Cao, Fei Shi, Xinjian Chen, Jili Chen
Summary: AI-based screening for diabetic retinopathy and macular edema using fundus photos and OCT images shows high sensitivity and specificity in a community hospital.
BIOMEDICAL ENGINEERING ONLINE
(2022)
Article
Computer Science, Artificial Intelligence
Mamta Juneja, Janmejai Singh Minhas, Naveen Singla, Sarthak Thakur, Niharika Thakur, Prashant Jindal
Summary: This research investigates a fused framework for assisting in the diagnosis of glaucoma using machine learning and deep learning techniques, presenting a method of fusing predicted classes through Major Voting and Weighted Decision Fusion. Experimental results demonstrate that this framework outperforms deep learning approaches based on other models, with a Precision of 0.95, Recall of 0.97 and F1-Score of 0.96, while also providing Grad-CAMs to aid medical experts in diagnosis.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Ophthalmology
Pengxiao Zang, Tristan T. Hormel, Xiaogang Wang, Kotaro Tsuboi, David Huang, Thomas S. Hwang, Yali Jia
Summary: The study evaluated a deep learning framework for DR classification using OCT and OCTA, showing high accuracy in classification.
TRANSLATIONAL VISION SCIENCE & TECHNOLOGY
(2022)
Article
Computer Science, Information Systems
Qaisar Abbas, Mostafa E. A. Ibrahim, Abdul Rauf Baig
Summary: This paper proposes a computer-aided diagnosis (CAD) system for diagnosing diabetic retinopathy (DR). The system uses preprocessing and a pre-train transfer learning-based convolutional neural network (PCNN) to recognize the five stages of DR. The results demonstrate that the CAD-DR system outperforms other state-of-the-art methods in terms of sensitivity, specificity, and accuracy, making it suitable for DR screening.
CMC-COMPUTERS MATERIALS & CONTINUA
(2022)
Article
Medicine, General & Internal
Haifan Huang, Liangjiu Zhu, Weifang Zhu, Tian Lin, Leonoor Inge Los, Chenpu Yao, Xinjian Chen, Haoyu Chen
Summary: The study developed an algorithm using deep learning technology to detect and quantify HRDs on OCT for DME patients. The algorithm showed stronger correlation and higher ICC with rater 1 compared to inter-rater agreement, providing an objective and repeatable tool for OCT analysis in clinical practice and research.
FRONTIERS IN MEDICINE
(2021)
Article
Biochemical Research Methods
Xuehua Wang, Rui Li, Junyan Chen, Dingan Han, Mingyi Wang, Honglian Xiong, Wenzheng Ding, Yixu Zheng, Ke Xiong, Yaguang Zeng
Summary: A deep learning model called CVI-Net is proposed for automatic segmentation of the choroid layer and its vessels in OCT scans. It quantifies clinical parameters to determine structural and vascular changes in the choroid with the progression of DR severity. The results show high accuracy and a significant negative correlation between CVI and DR severity.
JOURNAL OF BIOPHOTONICS
(2023)
Article
Computer Science, Artificial Intelligence
Mohammad H. Alshayeji, Sa'ed Abed, Silpa ChandraBhasi Sindhu
Summary: Diabetic retinopathy (DR), a common cause of vision loss, lacks early symptoms, making diagnosis difficult. A reliable machine learning model is proposed for early diagnosis and disease-stage screening. The framework achieves high accuracy and identifies relevant features for accurate diagnosis.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Review
Medicine, General & Internal
Sara Vaz-Pereira, Tiago Morais-Sarmento, Michael Engelbert
Summary: This review discusses recent developments in the use of optical coherence tomography (OCT) for proliferative diabetic retinopathy (PDR), focusing on the higher detection rate of neovascularization and improving retinal imaging quality with widefield-OCT and OCT-angiography (OCTA). Additionally, studies have advanced in analyzing retinal nonperfusion areas (NPAs) and the impact of PDR treatment on NPAs and vascular density. Ongoing technological developments in artificial intelligence and deep learning are enhancing imaging protocols for managing PDR without the need for invasive conventional fluorescein angiography.
Article
Pharmacology & Pharmacy
Maria Consiglia Trotta, Carlo Gesualdo, Chiara Bianca Maria Platania, Domenico De Robertis, Mauro Giordano, Francesca Simonelli, Michele D'Amico, Filippo Drago, Claudio Bucolo, Settimio Rossi
Summary: This study identified five miRNAs associated with stages of diabetic retinopathy, validating them as potential prognostic biomarkers and pharmacological targets. Additionally, optical coherence tomography (OCT) assessment showed an increase in the number of hyperreflective spots (HRS) with worsening of DR stages.
BIOCHEMICAL PHARMACOLOGY
(2021)
Article
Biochemistry & Molecular Biology
Gianpaolo Zerbini, Silvia Maestroni, Ilaria Vigano, Andrea Mosca, Renata Paleari, Daniela Gabellini, Silvia Galbiati, Paolo Rama
Summary: Diabetes-driven retinal neurodegeneration is involved in the early stages of diabetic retinopathy, and identifying a biomarker for early retinal neurodegeneration is crucial. This study confirms that thinning of the retinal nerve fiber layer/ganglion cell layer (RNFL/GCL) precedes the death of retinal ganglion cells in a diabetic mouse model, suggesting it as a potential biomarker. Starting a neuroprotective treatment at the first sign of RNFL/GCL thinning not only prevents loss of these cells but also the development of microvascular diabetic retinopathy.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2023)
Article
Biochemical Research Methods
Chen Wang, Justin C. Reynolds, Paul Calle, Avery D. Ladymon, Feng Yan, Yuyang Yan, Sam Ton, Kar-ming Fung, Sanjay G. Patel, Zhongxin Yu, Chongle Pan, Qinggong Tang
Summary: In this study, a computer-aided endoscopic optical coherence tomography (OCT) system was developed to guide Veress needle insertion accurately. The system was tested using swine samples and showed high accuracy in tissue classification and distance estimation.
JOURNAL OF BIOPHOTONICS
(2022)
Article
Multidisciplinary Sciences
Zongyi Wang, Haiyan An, Jiyang Tang, Enzhong Jin, Siying Li, Linqi Zhang, Lvzhen Huang, Jinfeng Qu
Summary: This study quantitatively analyzed the number and density of macrophage-like cells (MLCs) at the vitreoretinal interface in diabetic retinopathy (DR) with and without diabetic macular edema (DME) using optical coherence tomography angiography (OCTA). The results showed that the number and density of MLCs were higher in the DME group compared to the non-DME group. In the non-DME group, NPDR eyes had a higher number and density of MLCs, greater central macular thickness, and vessel density compared to PDR eyes.
SCIENTIFIC REPORTS
(2023)
Article
Engineering, Biomedical
Amr Elsawy, Mohamed Abdel-Mottaleb
Summary: A new deep-learning network was proposed for the diagnosis of Fuchs' endothelial dystrophy and keratoconus based on OCT images. The network achieved higher accuracy in image and scan classification compared to other networks, indicating its potential for early diagnosis of corneal diseases.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Wenting Chen, Shuang Yu, Kai Ma, Wei Ji, Cheng Bian, Chunyan Chu, Linlin Shen, Yefeng Zheng
Summary: This paper proposes a novel Topology and Width Aware Generative Adversarial Network (TW-GAN) that integrates topology connectivity and vessel width information into the deep learning framework for automatic artery/vein (A/V) classification. Experimental results demonstrate that the proposed framework significantly improves the topological connectivity of segmented A/V masks and achieves state-of-the-art A/V classification performance on public datasets.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Computer Science, Artificial Intelligence
Yating Huang, Xuechen Li, Siting Zheng, Zhongliang Li, Sihan Li, Linlin Shen, Changen Zhou, Zhihui Lai
Summary: The size and shape of the tongue can reflect different pathological changes of the human body in Traditional Chinese Medicine (TCM). In this work, an efficient deep network, TSCWNet, is proposed for tongue size and shape classification. Experimental results demonstrate that the network achieves better classification performance for tongue diagnosis.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Ziqi Jin, Jinheng Xie, Bizhu Wu, Linlin Shen
Summary: In this paper, a weakly supervised pedestrian segmentation framework is proposed to directly generate the foreground mask from person re-identification datasets with only image-level subject ID labels. The Image Synthesis Augmentation (ISA) technique is also introduced to further enhance the dataset. Experimental results demonstrate that the proposed framework learns robust and discriminative features, achieving significant improvement in mAP compared to the baseline on widely used datasets including Market-1501, CUHK03, and MSMT17. The code will be made available soon.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Syed Furqan Qadri, Hongxiang Lin, Linlin Shen, Mubashir Ahmad, Salman Qadri, Salabat Khan, Maqbool Khan, Syeda Shamaila Zareen, Muhammad Azeem Akbar, Md Belal Bin Heyat, Saqib Qamar
Summary: This study proposes a patch-based deep learning approach for automatic CT vertebral segmentation. The method extracts discriminative features from unlabeled data using a stacked sparse autoencoder and achieves accurate segmentation of CT vertebrae.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Zhongliang Li, Xuechen Li, Zhihao Jin, Linlin Shen
Summary: In this paper, a novel self-supervised pretraining method based on pseudo-lesion generation and restoration was proposed for COVID-19 diagnosis. The method trained an encoder-decoder architecture-based U-Net using pairs of pseudo-COVID-19 images and normal CT images for image restoration, and then fine-tuned the pretrained encoder using labeled data. Experimental results demonstrated that the proposed method extracted better feature representation for COVID-19 diagnosis, achieving higher accuracy compared to the supervised model pretrained on large-scale images by 6.57% and 3.03% on SARS-CoV-2 dataset and Jinan COVID-19 dataset, respectively.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Jiajun Wen, Honglin Chu, Zhihui Lai, Tianyang Xu, Linlin Shen
Summary: This paper proposes an innovative method called EFSCF, which performs jointly sparse feature learning to handle the spatial boundary effect effectively while suppressing the influence of background pixels and noises. The proposed method achieves better tracking performance than the state-of-the-art trackers by exploring the structural sparsity in rows and columns of a learned filter simultaneously.
Article
Neurosciences
Xiuzhi Zhao, Wenting Chen, Weicheng Xie, Linlin Shen
Summary: This study proposes a Style Attention based Global-local Aware GAN to generate personalized caricatures. It integrates the facial characteristics of a subject through a landmark-based warp controller for personalized shape exaggeration and uses a style-attention module for appropriate fusion of facial features and caricature style. The results indicate that the proposed method can preserve the identity of input photos and generate caricatures close to those drawn by real artists.
FRONTIERS IN NEUROSCIENCE
(2023)
Article
Plant Sciences
Linlin Shen, Haiyan Deng, Ganglong Zhang, Anqi Ma, Xiaoyong Mo
Summary: Climate warming poses a significant threat to global ecosystems, impacting the geographic distribution and suitable growth areas of species. This study focused on predicting the potential cultivation regions of Castanopsis hystrix Miq., a research object, based on the MaxEnt model and environmental variables. The key factors affecting the distribution area of C. hystrix Miq. were identified as the minimum temperature of the coldest month, precipitation of the driest month, and precipitation of the warmest quarter. The suitable cultivation regions were found in central and southern China, with a range of 18-34°N and 89-122°E, covering an area of 261.95 x 10(4) km(2). Under different climate scenarios, the spatial pattern of C. hystrix Miq. will migrate to different regions, with varying changes in suitable area and cultivation distribution.
Article
Computer Science, Artificial Intelligence
Ya-Nan Zhang, Linlin Shen, Zhihui Lai
Summary: In the field of computer vision, removing rain streaks from images is an important task as it affects the performance and quality of subsequent tasks and outdoor images. Deep learning-based methods have been proposed to address this issue, but they lack interpretability and have limited performance in detail restoration and rain streak removal. This paper introduces a rain streaks model-driven deep network, MSANet, to overcome these limitations.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Automation & Control Systems
Junhong Zhang, Zhihui Lai, Heng Kong, Linlin Shen
Summary: In this paper, a new robust manifold twin bounded SVM (RMTBSVM) algorithm is proposed, which considers both robustness and discriminability. By using the capped L-1-norm as the distance metric and adding robust manifold regularization, the robustness and classification performance are improved. The algorithm is extended for nonlinear classification using the kernel method.
IEEE TRANSACTIONS ON CYBERNETICS
(2023)
Article
Computer Science, Artificial Intelligence
Siyang Song, Shashank Jaiswal, Enrique Sanchez, Georgios Tzimiropoulos, Linlin Shen, Michel Valstar
Summary: This article addresses two important issues in automatic personality analysis systems: the use of short video segments or single frames for inferring personality traits, and the lack of methods for encoding person-specific facial dynamics. To tackle these issues, the paper proposes a novel Rank Loss for self-supervised learning of facial dynamics and a method to represent person-specific dynamics. The approach achieves promising results in personality estimation and shows the importance of the tasks performed by the subject in the video and the use of multi-scale dynamics.
IEEE TRANSACTIONS ON AFFECTIVE COMPUTING
(2023)
Article
Thermodynamics
Lili Liu, Cai Chen, Linlin Shen, Gang Xu, Yufeng Wen, Xianshi Zing
Summary: The pressure dependence of lattice and elastic constants of g-TiAl, DO22-Al3Ti, and alpha(2)-Ti3Al binary precipitates was investigated using a first-principles approach. The calculated results at 0 GPa and 0 K were in good agreement with existing experimental and theoretical values. The temperature and pressure dependencies of bulk modulus, Gibbs free energy, thermal expansion coefficient, and heat capacity at constant pressure were systematically studied using density-functional perturbation theory (DFPT) under the quasiharmonic approximation (QHA) in the ranges of 0-1000 K and 0-30 GPa.
HIGH TEMPERATURES-HIGH PRESSURES
(2023)
Article
Computer Science, Artificial Intelligence
Yang Zhang, Linlin Shen
Summary: In this study, an automatic learning rate tuning method for memristive deep learning systems is presented. The method utilizes memristors to adjust the adaptive learning rate in deep neural networks. The proposed method is robust to noisy gradients, various architectures, and different datasets and can address the issue of over-fitting. Moreover, a quantized neural network architecture is utilized in the presented system, leading to an increase in training efficiency without the loss of testing accuracy.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Zhe Kong, Wentian Zhang, Feng Liu, Wenhan Luo, Haozhe Liu, Linlin Shen, Raghavendra Ramachandra
Summary: This study empirically proves the importance of model initialization for generalization and proposes a self-supervised learning-based method to address the issue of unknown PAI.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Proceedings Paper
Neuroimaging
Haoqin Ji, Haozhe Liu, Yuexiang Li, Jinheng Xie, Nanjun He, Yawen Huang, Dong Wei, Xinrong Chen, Linlin Shen, Yefeng Zheng
Summary: Accurate abnormality localization in chest X-rays (CXR) is important for clinical diagnosis, but obtaining lesion-level annotations is difficult and time-consuming. We propose a weakly semi-supervised strategy called Point Beyond Class (PBC) that utilizes fully annotated CXRs with bounding boxes and weakly annotated samples by points. PBC learns a mapping from point annotations to bounding boxes and utilizes regularization and self-supervision to improve accuracy. Experimental results show that PBC outperforms the state-of-the-art method in abnormality localization.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT III
(2022)
Article
Computer Science, Artificial Intelligence
Rui Lv, Dingheng Wang, Jiangbin Zheng, Zhao-Xu Yang
Summary: In this paper, the authors investigate tensor decomposition for neural network compression. They analyze the convergence and precision of tensor mapping theory, validate the rationality of tensor mapping and its superiority over traditional tensor approximation based on the Lottery Ticket Hypothesis. They propose an efficient method called 3D-KCPNet to compress 3D convolutional neural networks using the Kronecker canonical polyadic (KCP) tensor decomposition. Experimental results show that 3D-KCPNet achieves higher accuracy compared to the original baseline model and the corresponding tensor approximation model.
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Computer Science, Artificial Intelligence
Zhijian Zhuo, Bilian Chen, Shenbao Yu, Langcai Cao
Summary: In this paper, a novel method called Expansion with Contraction Method for Overlapping Community Detection (ECOCD) is proposed, which utilizes non-negative matrix factorization to obtain disjoint communities and applies expansion and contraction processes to adjust the degree of overlap. ECOCD is applicable to various networks with different properties and achieves high-quality overlapping community detection.
Article
Computer Science, Artificial Intelligence
Yizhe Zhu, Chunhui Zhang, Jialin Gao, Xin Sun, Zihan Rui, Xi Zhou
Summary: In this work, the authors propose a Contrastive Spatio-Temporal Distilling (CSTD) approach to improve the detection of high-compressed deepfake videos. The approach leverages spatial-frequency cues and temporal-contrastive alignment to fully exploit spatiotemporal inconsistency information.
Review
Computer Science, Artificial Intelligence
Laijin Meng, Xinghao Jiang, Tanfeng Sun
Summary: This paper provides a review of coverless steganographic algorithms, including the development process, known contributions, and general issues in image and video algorithms. It also discusses the security of coverless steganography from theoretical analysis to actual investigation for the first time.
Article
Computer Science, Artificial Intelligence
Yajie Bao, Tianwei Xing, Xun Chen
Summary: Visual question answering requires processing multi-modal information and effective reasoning. Neural-symbolic learning is a promising method, but current approaches lack uncertainty handling and can only provide a single answer. To address this, we propose a confidence based neural-symbolic approach that evaluates NN inferences and conducts reasoning based on confidence.
Article
Computer Science, Artificial Intelligence
Anh H. Vo, Bao T. Nguyen
Summary: Interior style classification is an interesting problem with potential applications in both commercial and academic domains. This project proposes a method named ISC-DeIT, which combines data-efficient image transformer architectures and knowledge distillation, to address the interior style classification problem. Experimental results demonstrate a significant improvement in predictive accuracy compared to other state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Shashank Kotyan, Danilo Vasconcellos Vargas
Summary: This article introduces a novel augmentation technique called Dynamic Scanning Augmentation to improve the accuracy and robustness of Vision Transformer (ViT). The technique leverages dynamic input sequences to adaptively focus on different patches, resulting in significant changes in ViT's attention mechanism. Experimental results demonstrate that Dynamic Scanning Augmentation outperforms ViT in terms of both robustness to adversarial attacks and accuracy against natural images.
Article
Computer Science, Artificial Intelligence
Hiba Alqasir, Damien Muselet, Christophe Ducottet
Summary: The article proposes a solution to improve the learning process of a classification network by providing shape priors, reducing the need for annotated data. The solution is tested on cross-domain digit classification tasks and a video surveillance application.
Article
Computer Science, Artificial Intelligence
Dexiu Ma, Mei Liu, Mingsheng Shang
Summary: This paper proposes a method using neural dynamics solvers to solve infinity-norm optimization problems. Two improved solvers are constructed and their effectiveness and superiority are demonstrated through theoretical analysis and simulation experiments.
Article
Computer Science, Artificial Intelligence
Francesco Gregoretti, Giovanni Pezzulo, Domenico Maisto
Summary: Active Inference is a computational framework that uses probabilistic inference and variational free energy minimization to describe perception, planning, and action. cpp-AIF is a header-only C++ library that provides a powerful tool for implementing Active Inference for Partially Observable Markov Decision Processes through multi-core computing. It is cross-platform and improves performance, memory management, and usability compared to existing software.
Article
Computer Science, Artificial Intelligence
Zelin Ying, Dawei Cheng, Cen Chen, Xiang Li, Peng Zhu, Yifeng Luo, Yuqi Liang
Summary: This paper proposes a novel stock market trends prediction framework called SMART, which includes a self-supervised stock technical data sequence embedding model S3E. By training with multiple self-supervised auxiliary tasks, the model encodes stock technical data sequences into embeddings and uses the learned sequence embeddings for predicting stock market trends. Extensive experiments on China A-Shares market and NASDAQ market prove the high effectiveness of our model in stock market trends prediction, and its effectiveness is further validated in real-world applications in a leading financial service provider in China.
Article
Computer Science, Artificial Intelligence
Hao Li, Hao Jiang, Dongsheng Ye, Qiang Wang, Liang Du, Yuanyuan Zeng, Liu Yuan, Yingxue Wang, C. Chen
Summary: DHGAT1, a dynamic hyperbolic graph attention network, utilizes hyperbolic metric properties to embed dynamic graphs. It employs a spatiotemporal self-attention mechanism and weighted node representations, resulting in excellent performance in link prediction tasks.
Article
Computer Science, Artificial Intelligence
Jiehui Huang, Zhenchao Tang, Xuedong He, Jun Zhou, Defeng Zhou, Calvin Yu-Chian Chen
Summary: This study proposes a progressive learning multi-scale feature blending model for image deraining tasks. The model utilizes detail dilation and texture extraction to improve the restoration of rainy images. Experimental results show that the model achieves near state-of-the-art performance in rain removal tasks and exhibits better rain removal realism.
Article
Computer Science, Artificial Intelligence
Lizhi Liu, Zilin Gao, Yinhe Wang, Yongfu Li
Summary: This paper proposes a novel discrete-time interconnected model for depicting complex dynamical networks. The model consists of nodes and edges subsystems, which consider the dynamic characteristic of both nodes and edges. By designing control strategies and coupling modes, the stabilization and synchronization of the network are achieved. Simulation results demonstrate the effectiveness of the proposed methods.