Article
Chemistry, Analytical
Jiajing Fan, Siqi Yang, Jiahao Liu, Zhen Zhu, Jianbiao Xiao, Liang Chang, Shuisheng Lin, Jun Zhou
Summary: This study proposes a high accuracy and ultra-low power ECG-derived respiration estimation processor for wearable sensor applications, which provides a more comfortable and affordable method for respiration monitoring.
Article
Engineering, Electrical & Electronic
Shuaicong Hu, Wenjie Cai, Tijie Gao, Mingjie Wang
Summary: The study proposes a hybrid transformer model based on self-attention mechanism for obstructive sleep apnea (OSA) detection using single-lead electrocardiogram. By introducing a new method for constructing raw inputs, the study achieved accurate detection of OSA.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Chemistry, Analytical
Xinqi Bao, Aime Kingwengwe Abdala, Ernest Nlandu Kamavuako
Summary: This study investigated the accuracy of ECG-derived respiratory rate, finding that electrode placement does not impact estimation, baseline wander and amplitude modulation algorithms perform best, and frequency domain features are helpful for accurate estimation of respiratory rate.
Article
Biology
Quanan Yang, Lang Zou, Keming Wei, Guanzheng Liu
Summary: This study proposed a new method for detecting obstructive sleep apnea (OSA) using a one-dimensional squeeze-and-excitation (SE) residual group network to thoroughly extract the complementary information between heart rate variability (HRV) and ECG-derived respiration (EDR). The method showed higher accuracy, sensitivity, and specificity compared to existing methods during testing.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Engineering, Biomedical
Yuxing Lin, Hongyi Zhang, Wanqing Wu, Xingen Gao, Fei Chao, Juqiang Lin
Summary: This study proposes an automatic sleep apnea classification model based on wavelet transform and neural networks. By converting the signal and introducing a cost-sensitive algorithm, it can accurately identify breathing events. The results show that the model achieves high classification accuracy and performance.
PHYSICAL AND ENGINEERING SCIENCES IN MEDICINE
(2023)
Article
Biology
Clara Garcia-Vicente, Gonzalo C. Gutierrez-Tobal, Jorge Jimenez-Garcia, Adrian Martin-Montero, David Gozal, Roberto Hornero
Summary: A novel deep-learning approach using raw electrocardiogram tracing (ECG) was proposed to simplify the diagnosis of pediatric OSA. A convolutional neural network (CNN) regression model was implemented to predict pediatric OSA and derive severity categories. The diagnostic performance of the CNN model outperformed previous algorithms relying on ECG-derived features. The proposed CNN model provides a simpler, faster, and more accessible diagnostic test for pediatric OSA.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Computer Science, Information Systems
M. Tanveer, S. Sharma, K. Muhammad
Summary: The proposed LS-LSTSVM addresses the shortcomings of TWSVM and LSTSVM by introducing a different Lagrangian function to eliminate the need for calculating inverse matrices, using the kernel trick directly for non-linear cases, and minimizing structural risk. These improvements aim to enhance classification accuracy on datasets, especially for large-scale problems.
ACM TRANSACTIONS ON INTERNET TECHNOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
Imran Razzak, Mohamed Reda Bouadjenek, Raghib Abu Saris, Weiping Ding
Summary: Traditional support vector machines (SVMs) are sensitive to outliers and corrupted data, leading to a deterioration in classification performance. This article proposes an efficient Support Matrix Machine that performs matrix recovery and feature selection simultaneously. It can handle high-dimensional data with corrupted columns and recover an intrinsic matrix of higher rank under incoherence and ambiguity conditions. The objective function combines matrix recovery, low rank, and joint sparsity, and the method leverages structural information and intrinsic data structure. Experimental results show significant improvements in BCI, face recognition, and person identification datasets, especially in the presence of outliers.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Shi Cheng, JinBao Zhang, Zhan Gao, Jiehua Wang
Summary: This study presents the design of an integrated circuit for ECG-derived respiration using the wavelet transform method. The algorithm based on discrete wavelet transform was validated in Matlab and a corresponding digital circuit was designed. Experimental results show successful extraction of respiratory information from ECG data using the circuit.
JOURNAL OF DATABASE MANAGEMENT
(2022)
Article
Physics, Multidisciplinary
Duan Liang, Shan Wu, Lan Tang, Kaicheng Feng, Guanzheng Liu
Summary: This study found that NPSampEn is a more effective index for distinguishing patients with obstructive sleep apnea (OSA), demonstrating higher screening accuracy and stronger association with the apnea-hypopnea index (AHI) compared to other indices like LF/HF and SampEn.
Article
Computer Science, Artificial Intelligence
Pei-Yi Hao, Jung-Hsien Chiang, Yu-De Chen
Summary: This paper proposes a novel possibilistic classification algorithm using support vector machines (SVMs) to effectively handle uncertain information and improve classification performance. The algorithm aims at finding a maximal-margin fuzzy hyperplane based on possibility theory and solves a fuzzy mathematical optimization problem. The proposed algorithm retains the advantages of fuzzy set theory and SVM theory, and it is more robust for handling outliers. Experimental results demonstrate the satisfactory generalization accuracy and ability to describe inherent vagueness in the given dataset.
Article
Automation & Control Systems
Guanjin Wang, Kup-Sze Choi, Jeremy Yuen-Chun Teoh, Jie Lu
Summary: This article introduces a new approach called DCOT-LS-SVMs, which is based on least-squares support vector machines and utilizes deep cross-output knowledge transfer. The approach improves the generalizability of LS-SVMs and simplifies the parameter tuning process. Experimental results on UCI datasets and a prostate cancer diagnosis case study demonstrate the effectiveness of the proposed approach.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Medicine, General & Internal
Mohamed Sraitih, Younes Jabrane, Amir Hajjam El Hassani
Summary: The article aims to design an automatic arrhythmia classification system across patients, using a new ECG database segmentation paradigm that does not require feature extraction to improve arrhythmia detection. Experimental results show that in terms of computational cost, the SVM classifier outperforms other methods, making it suitable for clinical ECG classification models.
JOURNAL OF CLINICAL MEDICINE
(2021)
Article
Engineering, Biomedical
Hang Liu, Shaowei Cui, Xiaohui Zhao, Fengyu Cong
Summary: This paper proposes a CNN-Transformer architecture for automatic detection of obstructive sleep apnea (OSA) based on single-channel electrocardiogram (ECG) signals. The proposed method effectively improves the classification performance with a per-segment accuracy of 88.2% and an area under the receiver operating characteristic curve (AUC) of 0.95. It provides a promising and reliable solution for home portable detection of OSA.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Engineering, Electrical & Electronic
D. S. Parihar, Ripul Ghosh, Aparna Akula, Satish Kumar, H. K. Sardana
Summary: This article presents a hybrid approach using variational mode decomposition and feature extraction to classify elephant movements in a forest environment. The proposed method improves the accuracy of classification for elephants and other movements through the use of support vector machines.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Psychology, Multidisciplinary
Dunia J. Mahboobeh, Sofia B. Dias, Ahsan H. Khandoker, Leontios J. Hadjileontiadis
Summary: This study explores the use of ICT-based tools for capturing the status of patients with Parkinson's Disease (PD). By utilizing the Personalized Serious Game Suite and intelligent Motor Assessment Tests, the study found that high classification accuracy can be achieved from these data sources, effectively reflecting the motor skill status of PD patients and using machine learning to infer the stage of the disease. This integrated approach provides new opportunities for remote monitoring of PD patients' stages and contributes to more efficient organization and personalized interventions.
FRONTIERS IN PSYCHOLOGY
(2022)
Article
Engineering, Biomedical
Namareq Widatalla, Kiyoe Funamoto, Motoyoshi Kawataki, Chihiro Yoshida, Kenichi Funamoto, Masatoshi Saito, Yoshiyuki Kasahara, Ahsan Khandoker, Yoshitaka Kimura
Summary: This study used a mathematical model to estimate the QT intervals in fetal mice and validated the results with Doppler ultrasound measurements, showing good agreement between the two.
BIOMEDICAL ENGINEERING ONLINE
(2022)
Article
Health Care Sciences & Services
Peter Lee, Heepyung Kim, Yongshin Kim, Woohyeok Choi, M. Sami Zitouni, Ahsan Khandoker, Herbert F. Jelinek, Leontios Hadjileontiadis, Uichin Lee, Yong Jeong
Summary: This paper reviews smart masks that have emerged after the pandemic and explores their expansion, sensor technologies, and application platforms. Smart masks can address breathing discomfort from prolonged use and can be used for sensing COVID-19 and general health monitoring. Additionally, smart masks can enable group or community sensing, increasing the range and reliability of information. The service application fields for smart masks include daily-life health monitoring, sports training, protection for industry workers and soldiers, as well as respiratory hygiene in emergency rooms and ambulatory settings. Design considerations include sensor reliability, ergonomic design for user acceptance, and privacy-aware data handling.
JMIR MHEALTH AND UHEALTH
(2022)
Article
Cardiac & Cardiovascular Systems
Mohanad Alkhodari, Namareq Widatalla, Maisam Wahbah, Raghad Al Sakaji, Kiyoe Funamoto, Anita Krishnan, Yoshitaka Kimura, Ahsan H. Khandoker
Summary: This study proposes a novel artificial intelligence approach, deep coherence, which uses non-invasive electrocardiography to explain the relationship between maternal and fetal heartbeats during pregnancy. The performance of this approach was validated using a deep learning tool, achieving high accuracy in identifying coupling scenarios. The interpretability of deep learning was significant in explaining synchronization mechanisms between the maternal and fetal heartbeats.
FRONTIERS IN CARDIOVASCULAR MEDICINE
(2022)
Article
Computer Science, Information Systems
M. Sami Zitouni, Cheul Young Park, Uichin Lee, Leontios J. Hadjileontiadis, Ahsan Khandoker
Summary: This paper presents a framework for emotion recognition based on multi-modal peripheral signals, which can be implemented in daily life settings. The study collected emotion data from a debate using wearable devices and converted the emotions into classes in the arousal and valence space. The proposed framework achieved classification accuracy of > 96% and > 93% for independent and combined classification scenarios, respectively.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Information Systems
Ahsan Habib, Chandan Karmakar, John Yearwood
Summary: Deep-learning-based QRS-detection algorithms require post-processing to enhance R-peak localization accuracy. These algorithms use basic signal-processing tasks such as noise removal with a Salt and Pepper filter and domain-specific thresholds. However, the variation of these thresholds among studies and their bias towards the training data may result in performance drop in unknown test datasets. This study introduces a domain-agnostic automated post-processing using a separate recurrent neural network (RNN)-based model, which shows superiority over domain-specific post-processing in most cases.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Review
Psychology, Multidisciplinary
Nayeefa Chowdhury, Ahsan H. Khandoker
Summary: A literature review suggests that virtual reality exposure therapy (VRET) is as effective as in vivo exposure therapy (ET) for social anxiety disorder (SAD), but behavioral therapy based on classical conditioning principles has higher attrition and relapse rates. Further research is needed to compare the efficacy of the Pavlovian extinction model with other treatment models and to include neural markers as efficacy measures for treating SAD. A paradigm shift in the gold-standard treatment for SAD requires rigorous longitudinal comparative studies.
FRONTIERS IN PSYCHOLOGY
(2023)
Article
Engineering, Biomedical
Ahsan Habib, Mohammod Abdul Motin, Thomas Penzel, Marimuthu Palaniswami, John Yearwood, Chandan Karmakar
Summary: In this study, an automatic technique for multi-stage sleep classification using photoplethysmographic (PPG) signal is proposed. A convolutional neural network (CNN) is used to directly learn from the PPG signal and classify multiple sleep stages. Experimental results show that the proposed CNN model outperforms existing state-of-the-art models (classical and deep learning) in the literature.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING
(2023)
Article
Multidisciplinary Sciences
Shiza Saleem, Ahsan H. Khandoker, Mohanad Alkhodari, Leontios J. Hadjileontiadis, Herbert F. Jelinek
Summary: Heart failure is characterized by abnormal autonomic modulation, with sympathetic activation and parasympathetic withdrawal. Beta-blockers can inhibit sympathetic overstimulation and are used for heart failure patients with reduced ejection fraction. The effect of beta-blocker therapy on heart failure with preserved ejection fraction (HFpEF) is uncertain. In this study, ECGs of 73 HFpEF patients were analyzed to evaluate the impact of beta-blockers on heart rate variability (HRV) measures associated with cardiac risk.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Information Systems
Ahsan Habib, Chandan Karmakar, John Yearwood
Summary: Interpretability is crucial for achieving transparency and trust in machine learning models, especially in sensitive contexts like healthcare. This article explores the idea of explaining a model's decision-making by observing the sinc-kernels of a convolutional neural network. The optimized frequency-bands of these kernels can provide domain-specific insights, which are visualized through explanation vectors to identify significant frequency-bands for interpretation. The study also demonstrates the effectiveness of optimizing a CNN model using a subset of prominent sinc frequency-bands for task-specific interpretability.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Editorial Material
Physiology
Ahsan H. Khandoker, Ryoichi Nagatomi, Janos Negyesi
FRONTIERS IN PHYSIOLOGY
(2023)
Article
Multidisciplinary Sciences
Mohammod Abdul Motin, Chandan Karmakar, Marimuthu Palaniswami, Thomas Penzel, Dinesh Kumar
Summary: The traditional approach to monitoring sleep stages is inconvenient and requires expert support. We propose a single-sensor PPG-based automated sleep classification method. Experimental study with 10 patients showed an overall accuracy of 84.66%, 79.62%, and 72.23% for two-, three-, and four-stage sleep classification using SVM with polynomial kernel. These findings open opportunities for PPG-based wearable solutions for home-based automated sleep monitoring.
ROYAL SOCIETY OPEN SCIENCE
(2023)
Review
Cardiac & Cardiovascular Systems
Mohanad Alkhodari, Zhaohan Xiong, Ahsan H. Khandoker, Leontios J. Hadjileontiadis, Paul Leeson, Winok Lapidaire
Summary: This review discusses the integration of artificial intelligence (AI) and big data analysis for personalized cardiovascular care, specifically in the management of hypertensive disorders of pregnancy (HDP). The use of AI can provide personalized recommendations based on a deeper analysis of medical history and imaging data, leading to improved knowledge on pregnancy-related disorders and personalized treatment planning.
EXPERT REVIEW OF CARDIOVASCULAR THERAPY
(2023)
Article
Computer Science, Information Systems
Murad Almadani, Leontios Hadjileontiadis, Ahsan Khandoker
Summary: Fetal cardiac monitoring is crucial for early detection of potential fetal cardiac abnormalities, enabling prompt preventative care and safe births.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Information Systems
Ayman Youssef, Mohamed Abdelrazek, Chandan Karmakar
Summary: This paper presents machine learning models based on different ensemble algorithms for detecting software exploitation using runtime traces. The models were tested on 11 Windows applications, achieving up to 100% recall with 0% false positive rate. The study also discusses the impact of parameters for curating runtime traces on the performance of the models, and highlights important features and key takeaways.