Article
Health Care Sciences & Services
Takanobu Hirosawa, Takahiro Ito, Yukinori Harada, Kohei Ikenoya, Masashi Yokose, Taro Shimizu
Summary: This study evaluated the effect of phonocardiograms on the diagnostic accuracy in remote auscultation. The results showed that using a phonocardiogram improved the overall correct rate of remote auscultation by more than 10% and helped physicians differentiate abnormal heart sounds from normal ones.
Article
Veterinary Sciences
T. Vezzosi
Summary: This study evaluated a new smartphone-based digital stethoscope that can simultaneously record phonocardiographic and one-lead electrocardiogram (ECG) in dogs and cats. The recordings obtained were compared with conventional auscultation and standard ECG. The results showed that the new device demonstrated good diagnostic accuracy in detecting heart murmurs, gallop sounds, and cardiac arrhythmias.
VETERINARY JOURNAL
(2023)
Article
Automation & Control Systems
Ihsan Topaloglu, Prabal Datta Barua, Arif Metehan Yildiz, Tugce Keles, Sengul Dogan, Mehmet Baygin, Huseyin Fatih Gul, Turker Tuncer, Ru-San Tan, U. Rajendra Acharya
Summary: This study proposes an accurate asthma detection model using an attention network and machine learning technique. The model achieved a high accuracy of 99.73% in detecting asthma, surpassing previous models. The study also introduced a novel deep feature engineering model that can effectively distinguish asthma lung sounds from those of normal individuals.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Information Systems
Keying Ma, Jianbo Lu, Benzhuo Lu
Summary: In this article, a parameter-efficient densely connected dual attention network (DDA) for heart sound classification is proposed, which combines the advantages of end-to-end architecture and enriched contextual representations of the self-attention mechanism. Extensive experiments across stratified 10-fold cross-validation demonstrate that the DDA model surpasses current 1D deep models on the challenging Cinc2016 benchmark with significant computational efficiency.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Computer Science, Information Systems
Samiul Based Shuvo, Shams Nafisa Ali, Soham Irtiza Swapnil, Mabrook S. Al-Rakhami, Abdu Gumaei
Summary: The proposed CardioXNet model shows outstanding performance in automatic detection of five classes of cardiac auscultation, outperforming previous methods by achieving up to 99.60% accuracy, 99.56% precision, 99.52% recall and 99.68% F1-score on average. The model has been tested on multiple datasets and demonstrated high accuracy metrics, making it suitable for CVD screening in low resource setups using memory constraint mobile devices.
Article
Computer Science, Information Systems
Andoni Elola, Elisabete Aramendi, Jorge Oliveira, Francesco Renna, Miguel T. T. Coimbra, Matthew A. A. Reyna, Reza Sameni, Gari D. D. Clifford, Ali Bahrami Rad
Summary: This study aims to estimate the murmur grade of pediatric patients from a low-resource rural area using cardiac auscultation. The proposed method uses convolutional residual neural networks and attention mechanisms to classify phonocardiograms, and achieves high accuracy and performance in cross-validation and hidden test set.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Jay Karhade, Shaswati Dash, Samit Kumar Ghosh, Dinesh Kumar Dash, Rajesh Kumar Tripathy
Summary: This article proposes a time-frequency-domain deep learning framework for the automatic detection of heart valve disorders (HVDs) using phonocardiogram (PCG) signals. The proposed approach achieves high overall accuracy values in detecting HVDs and can classify normal and abnormal heart sound classes with good accuracy.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Article
Computer Science, Information Systems
Batyrkhan Omarov, Nurbek Saparkhojayev, Shyrynkyz Shekerbekova, Oxana Akhmetova, Meruert Sakypbekova, Guldina Kamalova, Zhanna Alimzhanova, Lyailya Tukenova, Zhadyra Akanova
Summary: Cardiovascular diseases are a leading cause of death globally. This study presents a prototype of a digital stethoscopic system that uses machine learning methods to diagnose cardiac abnormalities in real time. The system achieves an accuracy of over 90% in identifying abnormal heart sounds and offers the key advantage of speed and convenience.
CMC-COMPUTERS MATERIALS & CONTINUA
(2022)
Article
Medical Informatics
Chuan Yang, Wei Zhang, Zhixuan Pang, Jing Zhang, Deling Zou, Xinzhong Zhang, Sicong Guo, Jiye Wan, Ke Wang, Wenyue Pang
Summary: This study developed a low-cost electronic stethoscope, Auscul Pi, which enables ear-contactless auscultation and was clinically assessed against a standard electronic stethoscope. Results demonstrated that Auscul Pi showed similar real-time recording and playback performance to the standard stethoscope, with consistent data results such as phonocardiograms.
JMIR MEDICAL INFORMATICS
(2021)
Article
Engineering, Biomedical
Ximing Liao, Yin Wu, Nana Jiang, Jiaxing Sun, Wujian Xu, Shaoyong Gao, Jun Wang, Ting Li, Kun Wang, Qiang Li
Summary: This study developed a new Chinese adult respiratory sound database and used a lightweight neural network architecture for automated categorization of respiratory sounds. The study demonstrates the feasibility and potential of using mobile phones and electronic stethoscopes in computerized respiratory sound analysis, offering a convenient and promising approach to enhance respiratory disease management.
BIOCYBERNETICS AND BIOMEDICAL ENGINEERING
(2023)
Article
Medical Informatics
Wanrong Yang, Jiajie Xu, Junhong Xiang, Zhonghong Yan, Hengyu Zhou, Binbin Wen, Hai Kong, Rui Zhu, Wang Li
Summary: The diagnosis of cardiac abnormalities based on heart sound signal is a research hotspot. This study proposed a novel fuzzy matching feature extraction method for the automatic recognition of cardiac abnormalities. Machine learning algorithms were used to recognize heart disease based on the extracted features, and the results showed high classification accuracy. The combination of features achieved the best performance in support vector machine.
BMC MEDICAL INFORMATICS AND DECISION MAKING
(2022)
Article
Computer Science, Information Systems
Theekshana Dissanayake, Tharindu Fernando, Simon Denman, Sridha Sridharan, Houman Ghaemmaghami, Clinton Fookes
Summary: Traditional abnormal heart sound classification involves a three-stage process, with debate over whether to segment heart sounds before feature extraction. This study examines the importance of heart sound segmentation and proposes a robust, explainable classifier with nearly 100% accuracy on the PhysioNet dataset.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Computer Science, Interdisciplinary Applications
Kun Hu, Wenhua Wu, Wei Li, Milena Simic, Albert Zomaya, Zhiyong Wang
Summary: A novel deep learning architecture, A-ENN, is proposed for longitudinal grading of knee osteoarthritis (KOA) severity. By obtaining evolution traces through an adversarial training scheme, the fine-grained domain knowledge is fused with general convolutional image representations, achieving longitudinal grading.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2022)
Article
Computer Science, Interdisciplinary Applications
Juwairiya Siraj Khan, Manoj Kaushik, Anushka Chaurasia, Malay Kishore Dutta, Radim Burget
Summary: This paper proposes a deep learning method for automated diagnosis of cardiac diseases, achieving high accuracy through the combination of CNN and Cardi-Net systems to extract features from PCG signals. The model is robust and reliable for real-time applications, making it accessible in remote areas through cloud deployment and low-cost processors.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Computer Science, Information Systems
Lam Pham, Huy Phan, Ramaswamy Palaniappan, Alfred Mertins, Ian McLoughlin
Summary: This paper introduces a robust deep learning framework for analyzing respiratory sounds, aiming to classify anomalies and diseases. Through experiments on a respiratory sound dataset, different spectrogram types, spectral-time resolution, overlapping/non-overlapping windows, and data augmentation are explored for their impact on prediction accuracy. A novel deep learning system is proposed, outperforming current state-of-the-art methods, with a Teacher-Student scheme applied to balance model performance and complexity for potential real-time applications.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Computer Science, Information Systems
Ritesh Maurya, Neha Singh, Tanu Jindal, Vinay Kumar Pathak, Malay Kishore Dutta
Summary: This study investigated the effects of EMF radiation on fruit fly brain cells using transfer learning with CNN and SVM. The results showed that there was a significant impact on brain cells from fruit flies exposed to EMF radiation, supporting the initial hypothesis. The classification accuracy using pre-trained VGG19 network for feature extraction was 87.3% in a 5-fold cross-validation.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Ritesh Maurya, Vinay Kumar Pathak, Malay Kishore Dutta
Summary: This study introduces an automatic CapsNet framework for human epithelial cell image classification which addresses spatial relationships in IIF HEp-2 cell images better than CNNs, resulting in higher accuracy in identifying ANA-IIF images.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Engineering, Biomedical
Neeraj Baghel, Radim Burget, Malay Kishore Dutta
Summary: Fetal heart rate (FHR) is used as a screening tool to monitor the fetal state by obstetricians. Automated diagnostic technology based on artificial intelligence can assist in medical decisions and be used as an automatic diagnostic tool for primary health care centers and remote areas. A machine learning-based automated diagnostic tool using 1D-CNN model has been proposed to classify and diagnose Fetal Acidosis with high accuracy.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Article
Computer Science, Information Systems
Rakesh Chandra Joshi, Divyanshu Singh, Vaibhav Tiwari, Malay Kishore Dutta
Summary: The integration of deep learning technology with ultrasound images for pre-screening of breast cancer shows accurate and rapid performance in testing. The proposed method demonstrates high efficiency and effectiveness in real-time computer-aided diagnosis of breast tumors.
MULTIMEDIA TOOLS AND APPLICATIONS
(2022)
Article
Engineering, Chemical
Monika Arora, Parthasarathi Mangipudi, Malay Kishore Dutta
Summary: A novel mathematical model is proposed in this article to compute a distinct freshness coefficient, Q-score, by fusing information from relevant focal tissues of fish. Thorough investigation and normalization of multifocal tissues ensure the accuracy of feature extraction and integration into a single score. The implementation of the framework demonstrates an accuracy of 98.07%, indicating the efficacy of the proposed non-destructive method for rapid food quality evaluation.
JOURNAL OF FOOD ENGINEERING
(2022)
Article
Computer Science, Interdisciplinary Applications
Juwairiya Siraj Khan, Manoj Kaushik, Anushka Chaurasia, Malay Kishore Dutta, Radim Burget
Summary: This paper proposes a deep learning method for automated diagnosis of cardiac diseases, achieving high accuracy through the combination of CNN and Cardi-Net systems to extract features from PCG signals. The model is robust and reliable for real-time applications, making it accessible in remote areas through cloud deployment and low-cost processors.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Computer Science, Information Systems
Ritesh Maurya, Arti Srivastava, Ashutosh Srivastava, Vinay Kumar Pathak, Malay Kishore Dutta
Summary: Heavy metal pollution in aquatic bodies is a major health concern. Fishes are more susceptible to the harmful effects of non-biodegradable toxic metals. This study utilizes machine learning to identify heavy metal-contaminated fish using color and texture features, achieving high accuracy in classification. This technique can help mitigate the consequences of heavy metal toxicity and can be applied to large-scale fish processing.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Neha Sengar, Rakesh Chandra Joshi, Malay Kishore Dutta, Radim Burget
Summary: This paper presents an automated deep learning-based framework for diagnosing multiple eye diseases using colour fundus images. The EyeDeep-Net, a multi-layer neural network, is developed to extract relevant features from the input dataset and make predictive diagnostic decisions. The proposed model shows superior performance compared to baseline models in terms of classification and disease identification through digital fundus images.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Karnika Dwivedi, Malay Kishore Dutta
Summary: Blood-related diseases are a major concern in the biomedical field, and blood cell analysis is a key approach to diagnosis. This study proposes a CNN-based architecture called Microcell-Net, which is trained on a dataset of microscopic blood cell images in eight different classes. The experimental results demonstrate that the proposed model can efficiently classify various types of microscopic blood cells with good accuracy, outperforming other state-of-the-art models. The model's speed, automation, and efficiency make it suitable for real-time diagnosis and early detection of hematological disorders.
Article
Engineering, Civil
Rakesh Chandra Joshi, Dongryeol Ryu, Patrick N. J. Lane, Gary J. Sheridan
Summary: This study integrated remotely sensed plant response, meteorological forcing, and landscape attributes into a machine learning model to forecast summer soil moisture over forested landscapes. The results showed promising potential for applications in forest hydrology and bushfire risk planning.
JOURNAL OF HYDROLOGY
(2023)
Article
Medicine, General & Internal
Vojtech Myska, Samuel Genzor, Anzhelika Mezina, Radim Burget, Jan Mizera, Michal Stybnar, Martin Kolarik, Milan Sova, Malay Kishore Dutta
Summary: Pulmonary fibrosis is a severe long-term consequence of COVID-19, and corticosteroid treatment can increase the chances of recovery but may have side effects. This study developed prediction models using various algorithms to personalize the selection of patients benefiting from corticotherapy. The experiments proved the predictive value of information obtained during the initiation of post-COVID-19 treatment for personalized treatment decisions.
Proceedings Paper
Computer Science, Information Systems
Mohd Mohsin Ali, Rakesh Chandra Joshi, Malay Kishore Dutta, Radim Burget, Anzhelika Mezina
Summary: The challenge of categorizing viruses lies in their complex structures and the lighting conditions needed to capture TEM images. The proposed model demonstrated high accuracy on the validation and test sets, outperforming other deep-learning models. It is accurate, computationally less complex, and suitable for various medical applications in microscopic cell image analysis.
2022 45TH INTERNATIONAL CONFERENCE ON TELECOMMUNICATIONS AND SIGNAL PROCESSING, TSP
(2022)
Proceedings Paper
Computer Science, Information Systems
Arnav Kumar, Rakesh Chandra Joshi, Malay Kishore Dutta, Radim Burget, Vojtech Myska
Summary: Osteoporosis is a common global issue, and early diagnosis is crucial for proper treatment and reducing fractures. This study introduces a low-cost, efficient deep learning-based approach for diagnosing osteoporosis from X-ray images.
2022 45TH INTERNATIONAL CONFERENCE ON TELECOMMUNICATIONS AND SIGNAL PROCESSING, TSP
(2022)
Article
Multidisciplinary Sciences
Rakesh Chandra Joshi, Saumya Yadav, Malay Kishore Dutta, Carlos M. Travieso-Gonzalez
Summary: Blood cell analysis is crucial for health and immunity assessment. Traditional methods are time-consuming and expensive, necessitating the need for automated methods. This study proposes a convolutional neural network-based framework that can accurately detect and count various types of blood cells.