C‐TUnet: A CNN‐Transformer Architecture‐Based Ultrasound Breast Image Classification Network DOI Open Access
Ying Wu, Faming Li, Bo Xu

et al.

International Journal of Imaging Systems and Technology, Journal Year: 2024, Volume and Issue: 35(1)

Published: Dec. 17, 2024

ABSTRACT Ultrasound breast image classification plays a crucial role in the early detection of cancer, particularly differentiating benign from malignant lesions. Traditional methods face limitations feature extraction and global information capture, often resulting lower accuracy for complex noisy ultrasound images. This paper introduces novel network, C‐TUnet, which combines convolutional neural network (CNN) with Transformer architecture. In this model, CNN module initially extracts key features images, followed by module, captures context to enhance accuracy. Experimental results demonstrate that proposed model achieves excellent performance on public datasets, showing clear advantages over traditional methods. Our analysis confirms effectiveness combining modules—a strategy not only boosts robustness but also offers reliable tool clinical diagnostics, holding substantial potential real‐world application.

Language: Английский

Multi-threshold medical image segmentation based on the enhanced walrus optimizer DOI
Jie Li,

Ruicheng Lu,

Biqing Zeng

et al.

The Journal of Supercomputing, Journal Year: 2025, Volume and Issue: 81(4)

Published: Feb. 17, 2025

Language: Английский

Citations

0
A novel framework for efficient dominance-based rough set approximations using K-dimensional (K-D) tree partitioning and adaptive recalculations techniques DOI Creative Commons

Uzma Nawaz,

Zubair Saeed,

Kamran Atif

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 154, P. 110993 - 110993

Published: May 8, 2025

Language: Английский

Citations

0
Convolutional Neural Network Incorporating Multiple Attention Mechanisms for MRI Classification of Lumbar Spinal Stenosis DOI Creative Commons
Juncai Lin, Honglai Zhang, Hongcai Shang

et al.

Bioengineering, Journal Year: 2024, Volume and Issue: 11(10), P. 1021 - 1021

Published: Oct. 13, 2024

Lumbar spinal stenosis (LSS) is a common cause of low back pain, especially in the elderly, and accurate diagnosis critical for effective treatment. However, manual using MRI images time consuming subjective, leading to need automated methods.

Language: Английский

Citations

2
An Efficient Ensemble Approach for Brain Tumors Classification Using Magnetic Resonance Imaging DOI Creative Commons
Zubair Saeed, Tarraf Torfeh, Souha Aouadi

et al.

Information, Journal Year: 2024, Volume and Issue: 15(10), P. 641 - 641

Published: Oct. 15, 2024

Tumors in the brain can be life-threatening, making early and precise detection crucial for effective treatment improved patient outcomes. Deep learning (DL) techniques have shown significant potential automating diagnosis of tumors by analyzing magnetic resonance imaging (MRI), offering a more efficient accurate approach to classification. convolutional neural networks (DCNNs), which are sub-field DL, analyze rapidly accurately MRI data and, as such, assist human radiologists, facilitating quicker diagnoses earlier initiation. This study presents an ensemble three high-performing DCNN models, i.e., DenseNet169, EfficientNetB0, ResNet50, classification non-tumor samples. Our proposed model demonstrates improvements over various evaluation parameters compared individual state-of-the-art (SOTA) models. We implemented ten SOTA DenseNet121, SqueezeNet, ResNet34, ResNet18, VGG16, VGG19, LeNet5, provided detailed performance comparison. evaluated these models using two rates (LRs) 0.001 0.0001 batch sizes (BSs) 64 128 identified optimal hyperparameters each model. findings indicate that outperforms having 92% accuracy, 90% precision, recall, F1 score 91% at BS LR. not only highlights superior technique but also offers comprehensive comparison with latest research.

Language: Английский

Citations

1
Anterior Cruciate Ligament Tear Detection Based on T-Distribution Slice Attention Framework with Penalty Weight Loss Optimisation DOI Creative Commons
Weiqiang Liu, Yunfeng Wu

Bioengineering, Journal Year: 2024, Volume and Issue: 11(9), P. 880 - 880

Published: Aug. 30, 2024

Anterior cruciate ligament (ACL) plays an important role in stabilising the knee joint, prevents excessive anterior translation of tibia, and provides rotational stability. ACL injuries commonly occur as a result rapid deceleration, sudden change direction, or direct impact to during sports activities. Although several deep learning techniques have recently been applied detection tears, challenges such effective slice filtering nuanced relationship between varying tear grades still remain underexplored. This study used advanced model that integrated T-distribution-based attention mechanism with penalty weight loss function improve performance for tears. A T-distribution module was effectively utilised develop robust system model. By incorporating class relationships substituting conventional cross-entropy function, classification accuracy our is markedly increased. The combination shows significant improvements diagnostic across six different backbone networks. In particular, VGG-Slice-Weight provided area score 0.9590 under receiver operating characteristic curve (AUC). framework this offers tool supports better injury clinical diagnosis practice.

Language: Английский

Citations

0
C‐TUnet: A CNN‐Transformer Architecture‐Based Ultrasound Breast Image Classification Network DOI Open Access
Ying Wu, Faming Li, Bo Xu

et al.

International Journal of Imaging Systems and Technology, Journal Year: 2024, Volume and Issue: 35(1)

Published: Dec. 17, 2024

ABSTRACT Ultrasound breast image classification plays a crucial role in the early detection of cancer, particularly differentiating benign from malignant lesions. Traditional methods face limitations feature extraction and global information capture, often resulting lower accuracy for complex noisy ultrasound images. This paper introduces novel network, C‐TUnet, which combines convolutional neural network (CNN) with Transformer architecture. In this model, CNN module initially extracts key features images, followed by module, captures context to enhance accuracy. Experimental results demonstrate that proposed model achieves excellent performance on public datasets, showing clear advantages over traditional methods. Our analysis confirms effectiveness combining modules—a strategy not only boosts robustness but also offers reliable tool clinical diagnostics, holding substantial potential real‐world application.

Language: Английский

Citations

0