Two-Branch network for brain tumor segmentation using attention mechanism and super-resolution reconstruction

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 157, P. 106751 - 106751

Published: March 15, 2023

Language: Английский

---

Recent deep learning-based brain tumor segmentation models using multi-modality magnetic resonance imaging: a prospective survey

Frontiers in Bioengineering and Biotechnology, Journal Year: 2024, Volume and Issue: 12

Published: July 22, 2024

Radiologists encounter significant challenges when segmenting and determining brain tumors in patients because this information assists treatment planning. The utilization of artificial intelligence (AI), especially deep learning (DL), has emerged as a useful tool healthcare, aiding radiologists their diagnostic processes. This empowers to understand the biology better provide personalized care with tumors. segmentation using multi-modal magnetic resonance imaging (MRI) images received considerable attention. In survey, we first discuss available modalities properties. Subsequently, most recent DL-based models for tumor MRI. We divide section into three parts based on architecture: is that use backbone convolutional neural networks (CNN), second vision transformer-based models, third hybrid both transformer architecture. addition, in-depth statistical analysis performed publication, frequently used datasets, evaluation metrics tasks. Finally, open research are identified suggested promising future directions improve accuracy outcomes aligns public health goals technologies healthcare delivery population management.

Language: Английский

---

Advanced Medical Image Segmentation Enhancement: A Particle-Swarm-Optimization-Based Histogram Equalization Approach

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(2), P. 923 - 923

Published: Jan. 22, 2024

Accurate medical image segmentation is paramount for precise diagnosis and treatment in modern healthcare. This research presents a comprehensive study of the efficacy particle swarm optimization (PSO) combined with histogram equalization (HE) preprocessing segmentation, focusing on lung CT scan chest X-ray datasets. Best-cost values reveal PSO algorithm’s performance, HE demonstrating significant stabilization enhanced convergence, particularly complex images. Evaluation metrics, including accuracy, precision, recall, F1-score/Dice, specificity, Jaccard, show substantial improvements preprocessing, emphasizing its impact accuracy. Comparative analyses against alternative methods, such as Otsu, Watershed, K-means, confirm competitiveness PSO-HE approach, especially The also underscores positive influence clarity precision. These findings highlight promise approach advancing accuracy reliability pave way further method integration to enhance this critical healthcare application.

Language: Английский

---

An integrative machine learning model for the identification of tumor T-cell antigens

Biosystems, Journal Year: 2024, Volume and Issue: 237, P. 105177 - 105177

Published: March 1, 2024

Language: Английский

---

Augmented Transformer network for MRI brain tumor segmentation

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2024, Volume and Issue: 36(1), P. 101917 - 101917

Published: Jan. 1, 2024

The Augmented Transformer U-Net (AugTransU-Net) is proposed to address limitations in existing transformer-related models for brain tumor segmentation. While previous effectively capture long-range dependencies and global context, these works ignore the hierarchy a certain degree need more feature diversity as depth increases. AugTransU-Net integrates two advanced transformer modules into different positions within U-shaped architecture overcome issues. fundamental innovation lies constructing improved augmentation that incorporate Shortcuts standard blocks. These augmented are strategically placed at bottleneck of segmentation network, forming multi-head self-attention blocks circulant projections, aiming maintain enhance interaction diversity. Furthermore, paired attention operate from low high layers throughout establishing relationships both spatial channel dimensions. This allows each layer comprehend overall structure semantic information critical locations. Experimental results demonstrate effectiveness competitiveness comparison representative works. model achieves Dice values 89.7%/89.8%, 78.2%/78.6%, 80.4%/81.9% whole (WT), enhancing (ET) core (TC) on BraTS2019-2020 validation datasets, respectively. code will be made publicly available https://github.com/MuqinZ/AugTransUnet.

Language: Английский

---

Brain tumor segmentation based on deep learning, attention mechanisms, and energy-based uncertainty predictions

Multimedia Tools and Applications, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 20, 2025

Language: Английский

---

multiPI-TransBTS: A multi-path learning framework for brain tumor image segmentation based on multi-physical information

Computers in Biology and Medicine, Journal Year: 2025, Volume and Issue: 191, P. 110148 - 110148

Published: April 10, 2025

Language: Английский

---

DAUnet: A U-shaped network combining deep supervision and attention for brain tumor segmentation

Knowledge-Based Systems, Journal Year: 2023, Volume and Issue: 285, P. 111348 - 111348

Published: Dec. 27, 2023

Language: Английский

---

Performance Analysis of Segmentation and Classification of CT-Scanned Ovarian Tumours Using U-Net and Deep Convolutional Neural Networks

Diagnostics, Journal Year: 2023, Volume and Issue: 13(13), P. 2282 - 2282

Published: July 5, 2023

Difficulty in detecting tumours early stages is the major cause of mortalities patients, despite advancements treatment and research regarding ovarian cancer. Deep learning algorithms were applied to serve purpose as a diagnostic tool CT scan images region. The went through series pre-processing techniques and, further, tumour was segmented using UNet model. instances then classified into two categories—benign malignant tumours. Classification performed deep models like CNN, ResNet, DenseNet, Inception-ResNet, VGG16 Xception, along with machine such Random Forest, Gradient Boosting, AdaBoosting XGBoosting. DenseNet 121 emerges best model on this dataset after applying optimization by obtaining an accuracy 95.7%. current work demonstrates comparison multiple CNN architectures common algorithms, without applied.

Language: Английский

---

Automated Computer-Assisted Medical Decision-Making System Based on Morphological Shape and Skin Thickness Analysis for Asymmetry Detection in Mammographic Images

Diagnostics, Journal Year: 2023, Volume and Issue: 13(22), P. 3440 - 3440

Published: Nov. 14, 2023

Breast cancer is a significant health concern for women, emphasizing the need early detection. This research focuses on developing computer system asymmetry detection in mammographic images, employing two critical approaches: Dynamic Time Warping (DTW) shape analysis and Growing Seed Region (GSR) method breast skin segmentation. The methodology involves processing mammograms DICOM format. In morphological study, centroid-based mask computed using extracted images from files. Distances between centroid perimeter are then calculated to assess similarity through analysis. For thickness identification, seed initially set pixels expanded based intensity depth similarities. DTW achieves an accuracy of 83%, correctly identifying 23 possible cases out 20 ground truth cases. GRS validated Average Symmetric Surface Distance Relative Volumetric metrics, yielding similarities 90.47% 66.66%, respectively, compared 182 segmented successfully 35 patients with potential asymmetry. Additionally, Graphical User Interface designed facilitate insertion files provide visual representations asymmetrical findings validation accessibility by physicians.

Language: Английский

---