MARes-Net: multi-scale attention residual network for jaw cyst image segmentation DOI Creative Commons
Xiaokang Ding, Xiaoliang Jiang,

Huixia Zheng

и другие.

Frontiers in Bioengineering and Biotechnology, Год журнала: 2024, Номер 12

Опубликована: Авг. 5, 2024

Jaw cyst is a fluid-containing cystic lesion that can occur in any part of the jaw and cause facial swelling, dental lesions, fractures, other associated issues. Due to diversity complexity images, existing deep-learning methods still have challenges segmentation. To this end, we propose MARes-Net, an innovative multi-scale attentional residual network architecture. Firstly, connection used optimize encoder-decoder process, which effectively solves gradient disappearance problem improves training efficiency optimization ability. Secondly, scale-aware feature extraction module (SFEM) significantly enhances network's perceptual abilities by extending its receptive field across various scales, spaces, channel dimensions. Thirdly, compression excitation (MCEM) compresses excites map, combines it with contextual information obtain better model performance capabilities. Furthermore, introduction attention gate marks significant advancement refining map output. Finally, rigorous experimentation conducted on original dataset provided Quzhou People's Hospital verify validity MARes-Net The experimental data showed precision, recall, IoU F1-score reached 93.84%, 93.70%, 86.17%, 93.21%, respectively. Compared models, our shows unparalleled capabilities accurately delineating localizing anatomical structures image

Язык: Английский

Proton dose calculation on cone-beam computed tomography using unsupervised 3D deep learning networks DOI Creative Commons

Casper Dueholm Vestergaard,

U.V. Elstrøm,

L.P. Muren

и другие.

Physics and Imaging in Radiation Oncology, Год журнала: 2024, Номер 32, С. 100658 - 100658

Опубликована: Окт. 1, 2024

Язык: Английский

Процитировано

3

A systematic review of the role of artificial intelligence in automating computed tomography-based adaptive radiotherapy for head and neck cancer DOI Creative Commons
E. Mastella, Francesca Calderoni, Luigi Manco

и другие.

Physics and Imaging in Radiation Oncology, Год журнала: 2025, Номер 33, С. 100731 - 100731

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0

Quantitative use of cone-beam computed tomography in proton therapy: challenges and opportunities DOI Creative Commons
Hamid Ghaznavi, B. Maraghechi, H. Zhang

и другие.

Physics in Medicine and Biology, Год журнала: 2025, Номер 70(9), С. 09TR01 - 09TR01

Опубликована: Апрель 24, 2025

Abstract The fundamental goal in radiation therapy (RT) is to simultaneously maximize tumor cell killing and healthy tissue sparing. Reducing uncertainty margins improves normal sparing, but generally requires advanced techniques. Adaptive RT (ART) a compelling technique that leverages daily imaging anatomical information support reduced optimize plan quality for each treatment fraction. An especially exciting avenue ART proton (PT), which aims combine re-optimization with the unique advantages provided by protons, including integral dose near-zero deposition distal target along beam direction. A core component onboard image guidance, currently two options are available on systems, cone-beam computed tomography (CBCT) CT-on-rail (CToR) imaging. While CBCT suffers from poorer compared CToR imaging, platforms can be more easily integrated PT systems thus may streamlined adaptive (APT). In this review, we present current status of application evaluation adaptation, progress, challenges future directions.

Язык: Английский

Процитировано

0

A treatment-site-specific evaluation of commercial synthetic computed tomography solutions for proton therapy DOI Creative Commons
Ping Lin Yeap, Yun Ming Wong,

Kang Hao Lee

и другие.

Physics and Imaging in Radiation Oncology, Год журнала: 2024, Номер 31, С. 100639 - 100639

Опубликована: Июль 1, 2024

Despite the superior dose conformity of proton therapy, distribution is sensitive to daily anatomical changes, which can affect treatment accuracy. This study evaluated recalculation accuracy two synthetic computed tomography (sCT) generation algorithms in a commercial planning system.

Язык: Английский

Процитировано

2

НЕЙРОМЕРЕЖЕВИЙ ПІДХІД СЕГМЕНТАЦІЇ СІЛЬСЬКОГОСПОДАРСЬКИХ УГІДЬ НА СУПУТНИКОВИХ ЗОБРАЖЕННЯХ DOI Open Access

O. Honcharov,

Hnatushenko Vik.,

O. Shevtsova

и другие.

System technologies, Год журнала: 2024, Номер 4(153), С. 87 - 101

Опубликована: Май 1, 2024

Precision mapping and monitoring of agricultural lands using satellite imagery have become crucial for optimizing practices. This research focuses on ex-ploring the effectiveness deep learning models, particularly U-Net modifications, semantic segmentation in images. Recent Studies Publications Analysis. advancements convolutional neural networks (CNNs) shown promising results various tasks, including medical imaging, flood mapping, environmental monitoring. such as "Residual wave vision dual polarization Sentinel-1 SAR imagery" "Deep learning-based hybrid feature selection seg-mentation crops weeds" underline adaptability architectures to di-verse data characteristics, motivating their application land segmenta-tion. Research Objective. The primary aim this study is assess applicability efficiency modified accurately segmenting from It seeks identify optimal model modifications that enhance accuracy while maintaining computational efficiency, contributing more ef-fective Main Body Research. Utilizing images Copernicus HUB archive, work experiments with incorporating residual blocks, normalization methods, regularization techniques. compares perform-ance these models lands, highlighting impact archi-tectural enhancements improving precision generalization capabilities. Conclusions. concludes specific archi-tecture significantly Implementing batch normalization, dropout proved effective overcoming overfitting, suggesting a direction future geospatial processing agriculture. Further investigation into hyperparameter tuning, data-set expansion, ensemble methods recommended refine models' predictive performance.

Язык: Английский

Процитировано

0

MARes-Net: multi-scale attention residual network for jaw cyst image segmentation DOI Creative Commons
Xiaokang Ding, Xiaoliang Jiang,

Huixia Zheng

и другие.

Frontiers in Bioengineering and Biotechnology, Год журнала: 2024, Номер 12

Опубликована: Авг. 5, 2024

Jaw cyst is a fluid-containing cystic lesion that can occur in any part of the jaw and cause facial swelling, dental lesions, fractures, other associated issues. Due to diversity complexity images, existing deep-learning methods still have challenges segmentation. To this end, we propose MARes-Net, an innovative multi-scale attentional residual network architecture. Firstly, connection used optimize encoder-decoder process, which effectively solves gradient disappearance problem improves training efficiency optimization ability. Secondly, scale-aware feature extraction module (SFEM) significantly enhances network's perceptual abilities by extending its receptive field across various scales, spaces, channel dimensions. Thirdly, compression excitation (MCEM) compresses excites map, combines it with contextual information obtain better model performance capabilities. Furthermore, introduction attention gate marks significant advancement refining map output. Finally, rigorous experimentation conducted on original dataset provided Quzhou People's Hospital verify validity MARes-Net The experimental data showed precision, recall, IoU F1-score reached 93.84%, 93.70%, 86.17%, 93.21%, respectively. Compared models, our shows unparalleled capabilities accurately delineating localizing anatomical structures image

Язык: Английский

Процитировано

0