Review
Biochemical Research Methods
Yurui Chen, Louxin Zhang
Summary: This article introduces the application of deep learning in drug response prediction and summarizes the latest deep learning methods. Although deep learning methods have shown some limitations in certain cases, combining them with established bioinformatics analyses can help overcome some of these challenges.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Li Zhang, Heda Song, Nikolaos Aletras, Haiping Lu
Summary: Graph convolutional network (GCN) is an effective neural network model for graph representation learning. This paper proposes a new node-feature convolutional (NFC) layer to tackle the limitations of standard GCN. Experimental results show that NFC-GCN outperforms state-of-the-art methods in node classification.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Zonghan Wu, Da Zheng, Shirui Pan, Quan Gan, Guodong Long, George Karypis
Summary: This article introduces a novel spatial-temporal graph neural network called TraverseNet for capturing the spatial-temporal dependencies in traffic data. Compared to other spatial-temporal neural networks, TraverseNet views space and time as an inseparable whole and utilizes message traverse mechanisms to explore the dependencies in the spatial-temporal graph.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Fadi Dornaika, Jingjun Bi, Chongsheng Zhang
Summary: In recent years, semi-supervised learning on graphs has gained importance in many fields and applications. The goal is to use both labeled and unlabeled data to build more effective predictive models. However, existing graph convolutional networks (GCNs) have weaknesses in semi-supervised graph learning. To address this, the authors propose a novel semi-supervised learning method called ReNode-GLCNMR, which integrates graph learning and graph convolution into a unified network architecture and addresses the problem of topological imbalance.
Article
Engineering, Environmental
Ariele Zanfei, Andrea Menapace, Bruno M. Brentan, Robert Sitzenfrei, Manuel Herrera
Summary: This paper proposes a breakthrough approach for calibrating hydraulic models through a graph machine learning approach. The main idea is to create a graph neural network metamodel to estimate the network behaviour based on a limited number of monitoring sensors. Through this process, it is possible to estimate the uncertainty that is transferred from the few available measurements to the final hydraulic model.
Article
Computer Science, Artificial Intelligence
Tien Huu Do, Duc Minh Nguyen, Giannis Bekoulis, Adrian Munteanu, Nikos Deligiannis
Summary: Graph convolutional neural networks (GCNNs) have gained attention for their ability to handle graph-structured data, but face issues related to node transition probabilities, over-fitting, and over-smoothing. This study introduces a novel method to improve message passing based on transition probabilities and proposes DropNode regularization to address these challenges, demonstrating effectiveness in experiments.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Review
Computer Science, Information Systems
Manel Khazri Khlifi, Wadii Boulila, Imed Riadh Farah
Summary: This study investigates the recent developments of graph deep learning (GDL) in the field of remote sensing. It presents an extensive survey of the current state-of-the-art in GDL, with a specific focus on five established graph learning techniques. A taxonomy is proposed based on the input data type or task being considered, and several promising research directions are suggested to promote collaborations between the domains. This study is the first comprehensive review of graph deep learning in remote sensing.
COMPUTER SCIENCE REVIEW
(2023)
Article
Computer Science, Information Systems
Stefano De Sabbata, Pengyuan Liu
Summary: This study presents a systematic investigation on the use of graph neural networks for geodemographic classification. Using Greater London as a case study, the results show that the proposed Node Attributes-focused Graph AutoEncoder framework performs well in terms of class homogeneity and spatial clustering.
INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
(2023)
Article
Computer Science, Interdisciplinary Applications
Gangming Zhao, Kongming Liang, Chengwei Pan, Fandong Zhang, Xianpeng Wu, Xinyang Hu, Yizhou Yu
Summary: In this paper, a novel hybrid deep neural network for vessel segmentation is proposed. This network consists of two cascaded subnetworks for initial and refined segmentation, and utilizes cross-network multi-scale feature fusion to achieve high-quality vessel segmentation. The graph in the second subnetwork is constructed to tackle the challenges caused by vessel sparsity and anisotropy.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Computer Science, Information Systems
Hanyue Xu, Kah Phooi Seng, Li-Minn Ang
Summary: This paper proposes a hybrid approach that integrates graph convolutional networks (GCNs) and deep convolutional neural networks (DCNNs) for game strategies. Experimental results show that the hybrid model outperforms the traditional DCNN model in extracting game strategies.
Article
Computer Science, Artificial Intelligence
Tianyu Ma, Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
Summary: The convolutional neural network (CNN) is a commonly used architecture for computer vision tasks. A new building block called hyper-convolution is presented in this paper, which encodes the convolutional kernel using spatial coordinates and enables a more flexible architecture design. Experimental results showed that replacing regular convolutions with hyper-convolutions improved performance with fewer parameters and increased robustness against noise.
MEDICAL IMAGE ANALYSIS
(2023)
Article
Genetics & Heredity
Xuping Xie, Yan Wang, Nan Sheng, Shuangquan Zhang, Yangkun Cao, Yuan Fu
Summary: In this study, the authors propose a multi-view information fusion method called MVIFMDA for predicting miRNA-disease associations. The method combines multiple heterogeneous networks, graph convolutional networks, and attention strategies to effectively learn and fuse information from multi-source data. Experimental results demonstrate the superiority of the proposed model over baseline methods. The MVIFMDA model shows great potential in inferring underlying associations between miRNAs and diseases.
FRONTIERS IN GENETICS
(2022)
Article
Computer Science, Information Systems
Omar Del-Tejo-Catala, Jose-Luis Guardiola, Javier Perez, David Millan Escriva, Alberto J. Perez, Juan-Carlos Perez-Cortes
Summary: This paper proposes a novel probabilistic algorithm for pose estimation that addresses issues such as partial occlusion, object symmetries, and multiple correct poses. The algorithm combines information from multiple cameras to achieve accurate predictions. Testing on synthetic objects shows that the algorithm can handle these issues with a certain level of accuracy. Comparisons with state-of-the-art methodologies demonstrate that the algorithm can compete in terms of accuracy.
Article
Mathematics
Wenchuan Zhang, Weihua Ou, Weian Li, Jianping Gou, Wenjun Xiao, Bin Liu
Summary: Graph neural networks (GNNs) have gained attention for effectively processing graph-related data. Existing methods assume noise-free input graphs, which is frequently violated in real-world scenarios. To address this issue, we introduce virtual nodes and utilize Gumbel-Softmax to reweight edges, achieving differentiable graph structure learning (abbreviated as VN-GSL). Thorough evaluations on benchmark datasets demonstrate the superiority of our approach in terms of performance and efficiency. Our implementation will be publicly available.
Article
Computer Science, Artificial Intelligence
Zonghan Wu, Shirui Pan, Fengwen Chen, Guodong Long, Chengqi Zhang, Philip S. Yu
Summary: This article provides a comprehensive overview of graph neural networks (GNNs) in data mining and machine learning fields. It discusses the taxonomy of GNNs, their applications, and summarizes open-source codes, benchmark data sets, and model evaluation. The article also proposes potential research directions in this rapidly growing field.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2021)
Article
Computer Science, Interdisciplinary Applications
Zheng Chen, Naoaki Ono, Wei Chen, Toshiyo Tamura, Md Altaf-Ul-Amin, Shigehiko Kanaya, Ming Huang
Summary: The study introduced a method to predict malignant ventricular arrhythmias using signal complexity, which was validated through machine learning model experiments. This research provides important theoretical and practical implications for cardiac arrest prevention.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Editorial Material
Biology
Md. Altaf-Ul-Amin, Shigehiko Kanaya, Naoaki Ono, Ming Huang
Article
Biology
Koshiro Kido, Zheng Chen, Ming Huang, Toshiyo Tamura, Wei Chen, Naoaki Ono, Masachika Takeuchi, Md. Altaf-Ul-Amin, Shigehiko Kanaya
Summary: This study proposes a method for estimating blood pressure using PPG signal and evaluates its accuracy and robustness through the comparison of different regression models. The results show that an individual Gaussian Process model achieves the best performance, outperforming the generalized model built with all subjects' data.
Article
Biology
Xue Zhou, Keijiro Nakamura, Naohiko Sahara, Masako Asami, Yasutake Toyoda, Yoshinari Enomoto, Hidehiko Hara, Mahito Noro, Kaoru Sugi, Masao Moroi, Masato Nakamura, Ming Huang, Xin Zhu
Summary: This study utilized machine learning to identify three phenotypes of heart failure patients, stratifying them based on survival curves and mortality risk effectively. By training on the derivation dataset, these phenotypes were successfully applied to new patients in the validation dataset, with age and creatinine clearance rate identified as the top two most important predictors.
Article
Engineering, Biomedical
Shun Peng, Yang Li, Rui Cui, Ke Xu, Yonglin Wu, Ming Huang, Chenyun Dai, Toshiyo Tamur, Subhas Mukhopadhyay, Chen Chen, Wei Chen
Summary: This study investigates another potential application of cECG in sleep monitoring, specifically sleep posture recognition. By using a classifier model based on a deep recurrent neural network, accurate recognition of different sleep postures was achieved.
BIOMEDICAL ENGINEERING ONLINE
(2022)
Article
Infectious Diseases
Ahmad Kamal Nasution, Sony Hartono Wijaya, Pei Gao, Rumman Mahfujul Islam, Ming Huang, Naoaki Ono, Shigehiko Kanaya, Md Altaf-Ul-Amin
Summary: Jamu is a traditional Indonesian herbal medicine system that is considered to have many benefits. This study uses a machine learning approach to discover the potential of 14 plants as natural antibiotic candidates.
Article
Biochemical Research Methods
Zheng Chen, Ziwei Yang, Dong Wang, Xin Zhu, Naoaki Ono, M. D. Altaf-Ul-Amin, Shigehiko Kanaya, Ming Huang
Summary: Sleep screening is an important tool in healthcare and neuroscience research. Automatic sleep scoring using deep neural networks shows promising results, but lacks the medical criterion for consistent performance. This paper proposes a framework for sleep stage scoring that captures stage-specific features satisfying sleep medicine criteria. The framework includes feature extraction networks and an attention-based scoring decision network. The proposed method achieves competitive stage scoring performance, especially for Wake, N2, and N3 stages.
Article
Chemistry, Analytical
Toshiyo Tamura, Ming Huang, Takumi Yoshimura, Shinjiro Umezu, Toru Ogata
Summary: The study presents the design and prototype of an Internet of Things system for heatstroke prevention. It integrates physiological information, particularly deep body temperature (DBT), using the dual-heat-flux method. A dual-heat-flux thermometer was developed and evaluated for real-time DBT monitoring. Real-time readings are stored on a cloud platform and processed by a decision rule to alert users of heatstroke incidents.
Article
Physiology
Sota Kudo, Zheng Chen, Xue Zhou, Leighton T. Izu, Ye Chen-Izu, Xin Zhu, Toshiyo Tamura, Shigehiko Kanaya, Ming Huang
Summary: Photoplethysmography (PPG) signal shows potential in atrial fibrillation (AF) detection due to its convenience and physiological similarity to electrocardiogram (ECG). This study proposes a multiple-class classification model for AF detection, taking into consideration individual differences and sub-types in PPG manifestation. The best combination of configurable components in the pipeline includes first-order difference of heartbeat sequence as input format, a 2-layer CNN-1-layer Transformer hybrid model as the learning model, and the whole model fine-tuning as the transfer learning scheme (F1 value: 0.80, overall accuracy: 0.87).
FRONTIERS IN PHYSIOLOGY
(2023)
Article
Medicine, General & Internal
Keijiro Nakamura, Xue Zhou, Naohiko Sahara, Yasutake Toyoda, Yoshinari Enomoto, Hidehiko Hara, Mahito Noro, Kaoru Sugi, Ming Huang, Masao Moroi, Masato Nakamura, Xin Zhu
Summary: This study developed and validated a deep learning-based prognostic model to predict the risk of all-cause mortality for patients with HF. The proposed model showed better prediction performance in terms of discrimination, calibration, and risk stratification compared to other deep learning and traditional statistical models, especially in identifying high-risk patients.
Article
Engineering, Biomedical
Zheng Chen, Ziwei Yang, Lingwei Zhu, Wei Chen, Toshiyo Tamura, Naoaki Ono, Md Altaf-Ul-Amin, Shigehiko Kanaya, Ming Huang
Summary: In this paper, a novel framework is proposed for automated sleep staging based on sleep medicine guidance. The framework captures time-frequency characteristics of sleep EEG signals and utilizes a Transformer model with an attention-based module for staging decisions. The method achieves state-of-the-art results and demonstrates high inter-rater reliability, with important implications for healthcare and neuroscience research.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Proceedings Paper
Engineering, Biomedical
Ahmad Kamal Nasution, Sony Hartono Wijaya, Ming Huang, Naoaki Ono, Shigehiko Kanaya, Md. Altaf Ul-Amin
Summary: This research used machine learning methods to classify Jamu formulas and predict their effectiveness against bacterial diseases. It identified 111 potential antibiotic compounds for various systems.
2022 IEEE 22ND INTERNATIONAL CONFERENCE ON BIOINFORMATICS AND BIOENGINEERING (BIBE 2022)
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Guangxian Zhu, Huijia Wang, Yirong Kan, Zheng Chen, Ming Huang, Md. Amin, Naoaki Ono, Shigehiko Kanaya, Renyuan Zhang, Yasuhiko Nakashima
Summary: This paper presents an innovative non-deterministic coding method for EEG signals and achieves competitive results in sleep stage classification tasks.
2022 IEEE 35TH INTERNATIONAL SYSTEM-ON-CHIP CONFERENCE (IEEE SOCC 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Pei Gao, Zheng Chen, Ming Huang, Naoaki Ono, Shigehiko Kanaya, Md Altaf-UI-Amin
Summary: The study utilizes empirical data from Traditional Chinese Medicine to develop new antibiotics, screening out 2258 potential TCM formulae for treating bacterial pneumonia. Evaluated by the random forest algorithm, the matching labeling performs significantly better than clustering labeling by K-means.
2021 IEEE 3RD GLOBAL CONFERENCE ON LIFE SCIENCES AND TECHNOLOGIES (IEEE LIFETECH 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Guang Shi, Zhen Chen, Shigehiko Kanaya, Md Altaf-UI-Amin, Naoaki Ono, Ming Huang
Summary: In this study, machine learning algorithms are used to predict the body constitutions (BCs) of traditional Chinese medical theory. By identifying the principle features (PFs) of life-style, biased BCs are transformed into gentle constitutions to provide health guidance. The prediction accuracy is improved by 29% and the amount of identified PFs is reduced to 66.7% compared to previous works.
2021 IEEE 3RD GLOBAL CONFERENCE ON LIFE SCIENCES AND TECHNOLOGIES (IEEE LIFETECH 2021)
(2021)