Related references
Note: Only part of the references are listed.
Article
Radiology, Nuclear Medicine & Medical Imaging
Gurkan Kavuran et al.
Summary: This study aimed to develop and evaluate a fully automatic deep learning system for diagnosing COVID-19 using thoracic CT scans. The hybrid model MTU-COVNet showed an overall accuracy of 97.7% and high diagnostic performance for detecting COVID-19 and CAP, with specificity and sensitivity reaching 98.0% and 98.2% for COVID-19, and 99.1% and 97.1% for CAP.
Article
Computer Science, Artificial Intelligence
Serhat Kilicarslan et al.
Summary: Activation functions are crucial for neural network architectures to process input information and generate output. This study proposes extended versions of kernel-based activation functions that overcome existing issues and outperform other activation functions on benchmark datasets.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Mehmet Akif Bulbul et al.
Summary: This study develops an Artificial Neural Network (ANN)-based model to predict risk priorities for reinforced-concrete (RC) buildings in highly seismic-prone regions. The model achieves accurate predictions with maximum efficiency and a success rate of 98% in determining risk priorities.
APPLIED SCIENCES-BASEL
(2022)
Article
Chemistry, Analytical
Loris Nanni et al.
Summary: CNNs and other deep learners have become state-of-the-art in medical imaging research. However, the small sample size of many medical data sets hampers performance and leads to overfitting. Building deep CNN ensembles using pre-trained CNNs is a powerful method to overcome this problem. It relies on introducing diversity in the classification workflow, and a recent method is to vary activation functions in CNNs. This study aims to examine the performance of both methods using a large set of activation functions and demonstrates the superiority of this approach in various medical classification tasks.
Article
Computer Science, Information Systems
Cinare Oguz et al.
Summary: This study aims to reduce the duration and amount of COVID-19 transmission by shortening the diagnosis time of patients using Computed Tomography (CT). Deep learning models and classification methods were employed to develop a decision support system for radiologists. By extracting deep features and evaluating their performance, the study found that the combination of ResNet-50 and SVM achieved the best accuracy, F1-score, and AUC value. The high performance of this system suggests its potential as an auxiliary tool for diagnosing COVID-19.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Multidisciplinary Sciences
Abdullah Elen et al.
Summary: An adaptive kernel function based on the Gaussian kernel is designed in this study, which performs well in SVM and is compared with traditional linear, polynomial and Gaussian kernels.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2022)
Article
Computer Science, Interdisciplinary Applications
Kang Cheol Kim et al.
Summary: An automatic X-ray image segmentation technique combining deep-learning and level-set methods is proposed for compression fracture detection and evaluation. This structured hierarchical segmentation method utilizes pose-driven learning and M-net to accurately identify lumbar vertebrae and segment individual vertebrae. Fine tuning segmentation is achieved by combining the level-set method with the obtained segmentation results, resulting in accurate and robust identification of each lumbar vertebra.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Xinyu Liu et al.
Summary: The Tanh Exponential Activation Function (TanhExp) significantly improves the performance of lightweight or mobile neural networks in image classification tasks, displaying simplicity, efficiency, and robustness. Even with noise added and dataset altered, TanhExp's performance remains stable, enhancing network capacity without increasing the network size and requiring only a few training epochs with no extra parameters.
IET COMPUTER VISION
(2021)
Article
Computer Science, Artificial Intelligence
Hegui Zhu et al.
Summary: In this paper, a new nonlinear nonmonotonic activation function called Logish is proposed, demonstrating better performance in image classification tasks compared to other common activation functions.
Article
Radiology, Nuclear Medicine & Medical Imaging
Mohammad Salehi et al.
Summary: In this study, an automated convolutional neural network-based transfer learning approach was used to detect pneumonia in pediatric chest radiographs, with different pre-trained models evaluated for performance metrics. Results showed that all proposed models achieved accuracy greater than 83.0% for binary classification, with DenseNet121 model performing the best.
BRITISH JOURNAL OF RADIOLOGY
(2021)
Article
Biology
Ishak Pacal et al.
Summary: Colorectal cancer is the third most common type of cancer globally, and colonoscopy is the gold standard for screening. Deep learning architectures show promise in detecting polyps, but real-time detection success has not been achieved. By improving the YOLOv4 algorithm structure, the proposed method demonstrates superior performance on public datasets.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Serhat Kilicarslan et al.
Summary: In this study, novel RSigELU activation functions were proposed to address the issues of vanishing gradient and negative region problems, demonstrating better performance across various activation regions.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Interdisciplinary Applications
Gaobo Liang et al.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2020)
Article
Health Care Sciences & Services
Okeke Stephen et al.
JOURNAL OF HEALTHCARE ENGINEERING
(2019)
Article
Computer Science, Artificial Intelligence
Kemal Adem et al.
EXPERT SYSTEMS WITH APPLICATIONS
(2019)
Article
Computer Science, Artificial Intelligence
Kemal Adem et al.
TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES
(2019)
Article
Biochemistry & Molecular Biology
Daniel S. Kermany et al.
Article
Chemistry, Multidisciplinary
Vincent Menger et al.
APPLIED SCIENCES-BASEL
(2018)
Article
Multidisciplinary Sciences
Selcuk Bas
Review
Multidisciplinary Sciences
Yann LeCun et al.
Article
Immunology
Niranjan Bhat et al.
CLINICAL INFECTIOUS DISEASES
(2012)