A novel multi-head CNN design to identify plant diseases using the fusion of RGB images

Ecological Informatics, Год журнала: 2023, Номер 75, С. 101998 - 101998

Опубликована: Янв. 21, 2023

Язык: Английский

---

Automated detection and forecasting of COVID-19 using deep learning techniques: A review

Neurocomputing, Год журнала: 2024, Номер 577, С. 127317 - 127317

Опубликована: Янв. 26, 2024

Язык: Английский

---

Advanced Brain Tumor Classification in MR Images Using Transfer Learning and Pre-Trained Deep CNN Models

Cancers, Год журнала: 2025, Номер 17(1), С. 121 - 121

Опубликована: Янв. 2, 2025

Background/Objectives: Brain tumor classification is a crucial task in medical diagnostics, as early and accurate detection can significantly improve patient outcomes. This study investigates the effectiveness of pre-trained deep learning models classifying brain MRI images into four categories: Glioma, Meningioma, Pituitary, No Tumor, aiming to enhance diagnostic process through automation. Methods: A publicly available Tumor dataset containing 7023 was used this research. The employs state-of-the-art models, including Xception, MobileNetV2, InceptionV3, ResNet50, VGG16, DenseNet121, which are fine-tuned using transfer learning, combination with advanced preprocessing data augmentation techniques. Transfer applied fine-tune optimize accuracy while minimizing computational requirements, ensuring efficiency real-world applications. Results: Among tested Xception emerged top performer, achieving weighted 98.73% F1 score 95.29%, demonstrating exceptional generalization capabilities. These proved particularly effective addressing class imbalances delivering consistent performance across various evaluation metrics, thus their suitability for clinical adoption. However, challenges persist improving recall Glioma Meningioma categories, black-box nature requires further attention interpretability trust settings. Conclusions: findings underscore transformative potential imaging, offering pathway toward more reliable, scalable, efficient tools. Future research will focus on expanding diversity, model explainability, validating settings support widespread adoption AI-driven systems healthcare ensure integration workflows.

Язык: Английский

---

A survey on deep learning models for detection of COVID-19

Neural Computing and Applications, Год журнала: 2023, Номер 35(23), С. 16945 - 16973

Опубликована: Май 27, 2023

Язык: Английский

---

OPT-CO: Optimizing pre-trained transformer models for efficient COVID-19 classification with stochastic configuration networks

Information Sciences, Год журнала: 2024, Номер 680, С. 121141 - 121141

Опубликована: Июль 8, 2024

Building upon pre-trained ViT models, many advanced methods have achieved significant success in COVID-19 classification. Many scholars pursue better performance by increasing model complexity and parameters. While these can enhance performance, they also require extensive computational resources extended training times. Additionally, the persistent challenge of overfitting, due to limited dataset sizes, remains a hurdle. To address challenges, we proposed novel method optimize transformer models for efficient classification with stochastic configuration networks (SCNs), referred as OPT-CO. We two optimization methods: sequential (SeOp) parallel (PaOp), incorporating optimizers manner, respectively. Our without necessitating parameter expansion. introduced OPT-CO-SCN avoid overfitting problems through adoption random projection head augmentation. The experiments were carried out evaluate our based on publicly available datasets. Based evaluation results, superior, surpassing other state-of-the-art methods.

Язык: Английский

---

CNN-IKOA: convolutional neural network with improved Kepler optimization algorithm for image segmentation: experimental validation and numerical exploration

Journal Of Big Data, Год журнала: 2024, Номер 11(1)

Опубликована: Янв. 10, 2024

Abstract Chest diseases, especially COVID-19, have quickly spread throughout the world and caused many deaths. Finding a rapid accurate diagnostic tool was indispensable to combating these diseases. Therefore, scientists thought of combining chest X-ray (CXR) images with deep learning techniques rapidly detect people infected COVID-19 or any other disease. Image segmentation as preprocessing step has an essential role in improving performance techniques, it could separate most relevant features better train techniques. several approaches were proposed tackle image problem accurately. Among methods, multilevel thresholding-based methods won significant interest due their simplicity, accuracy, relatively low storage requirements. However, increasing threshold levels, traditional failed achieve segmented reasonable amount time. researchers recently used metaheuristic algorithms this problem, but existing still suffer from slow convergence speed stagnation into local minima number levels increases. study presents alternative technique based on enhanced version Kepler optimization algorithm (KOA), namely IKOA, segment CXR at small, medium, high levels. Ten are assess IKOA ten (T-5, T-7, T-8, T-10, T-12, T-15, T-18, T-20, T-25, T-30). To observe its effectiveness, is compared terms indicators. The experimental outcomes disclose superiority over all algorithms. Furthermore, IKOA-based eight different newly CNN model called CNN-IKOA find out effectiveness step. Five indicators, overall precision, recall, F1-score, specificity, CNN-IKOA’s effectiveness. CNN-IKOA, according outcomes, outstanding for where reach 94.88% 96.57% 95.40% recall.

Язык: Английский

---

Research on the multi-signal DOA estimation based on ResNet with the attention module combined with beamforming (RAB-DOA)

Applied Acoustics, Год журнала: 2025, Номер 231, С. 110541 - 110541

Опубликована: Янв. 21, 2025

Язык: Английский

---

An overview of deep learning techniques for COVID-19 detection: methods, challenges, and future works

Multimedia Systems, Год журнала: 2023, Номер 29(3), С. 1603 - 1627

Опубликована: Март 25, 2023

Язык: Английский

---

PSFHSP-Net: an efficient lightweight network for identifying pubic symphysis-fetal head standard plane from intrapartum ultrasound images

Medical & Biological Engineering & Computing, Год журнала: 2024, Номер 62(10), С. 2975 - 2986

Опубликована: Май 9, 2024

Abstract The accurate selection of the ultrasound plane for fetal head and pubic symphysis is critical precisely measuring angle progression. traditional method depends heavily on sonographers manually selecting imaging plane. This process not only time-intensive laborious but also prone to variability based clinicians’ expertise. Consequently, there a significant need an automated driven by artificial intelligence. To enhance efficiency accuracy identifying symphysis-fetal standard (PSFHSP), we proposed streamlined neural network, PSFHSP-Net, modified version ResNet-18. network comprises single convolutional layer three residual blocks designed mitigate noise interference bolster feature extraction capabilities. model’s adaptability was further refined expanding shared into task-specific layers. We assessed its performance against both heavyweight other lightweight models evaluating metrics such as F 1-score, (ACC), recall, precision, area under ROC curve (AUC), model parameter count, frames per second (FPS). PSFHSP-Net recorded ACC 0.8995, 1-score 0.9075, recall 0.9191, precision 0.9022. surpassed in these metrics. Notably, it featured smallest size (1.48 MB) highest processing speed (65.7909 FPS), meeting real-time criterion over 24 images second. While AUC our 0.930, slightly lower than that ResNet34 (0.935), showed marked improvement ResNet-18 testing, with increases 0.0435 0.0306, respectively. However, saw slight decrease from 0.9184 0.9022, reduction 0.0162. Despite trade-offs, compression significantly reduced 42.64 1.48 MB increased inference 4.4753 65.7909 FPS. results confirm capable swiftly effectively PSFHSP, thereby facilitating measurements development represents advancement automating analysis, promising enhanced consistency operator dependency clinical settings. Graphical abstract

Язык: Английский

---

Deep learning ensemble approach with explainable AI for lung and colon cancer classification using advanced hyperparameter tuning

BMC Medical Informatics and Decision Making, Год журнала: 2024, Номер 24(1)

Опубликована: Авг. 7, 2024

Lung and colon cancers are leading contributors to cancer-related fatalities globally, distinguished by unique histopathological traits discernible through medical imaging. Effective classification of these is critical for accurate diagnosis treatment. This study addresses challenges in the diagnostic imaging lung cancers, which among causes deaths worldwide. Recognizing limitations existing methods, often suffer from overfitting poor generalizability, our research introduces a novel deep learning framework that synergistically combines Xception MobileNet architectures. innovative ensemble model aims enhance feature extraction, improve robustness, reduce overfitting.Our methodology involves training hybrid on comprehensive dataset images, followed validation against balanced test set. The results demonstrate an impressive accuracy 99.44%, with perfect precision recall identifying certain cancerous non-cancerous tissues, marking significant improvement over traditional approach.The practical implications findings profound. By integrating Gradient-weighted Class Activation Mapping (Grad-CAM), offers enhanced interpretability, allowing clinicians visualize reasoning process. transparency vital clinical acceptance enables more personalized, treatment planning. Our not only pushes boundaries technology but also sets stage future aimed at expanding techniques other types cancer diagnostics.

Язык: Английский

---