FASNet: Feature Alignment-based method with digital pathology images in assisted diagnosis medical system DOI Creative Commons

Keke He,

Jun Zhu, Limiao Li

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(22), P. e40350 - e40350

Published: Nov. 1, 2024

Many important information in medical research and clinical diagnosis are obtained from images. Among them, digital pathology images can provide detailed tissue structure cellular information, which has become the gold standard for tumor diagnosis. With development of neural networks, computer-aided presents identification results various cell nuclei to doctors, facilitates cancerous regions. However, deep learning models require a large amount annotated data. Pathology expensive difficult obtain, insufficient annotation data easily lead biased results. In addition, when current evaluated on an unknown target domain, there errors predicted boundaries. Based this, this study proposes feature alignment-based detail recognition strategy image segmentation (FASNet). It consists preprocessing model network (UNW). The UNW performs instance normalization categorical whitening by inserting semantics-aware modules into encoder decoder, achieves compactness features same class separation different classes. FASNet method identify more efficiently, thus differentiate between classes tissues effectively. experimental show that Dice Similarity Coefficient (DSC) value 0.844. good performance even faced with test does not match distribution training Code: https://github.com/zlf010928/FASNet.git.

Language: Английский

Research on Artificial-Intelligence-Assisted Medicine: A Survey on Medical Artificial Intelligence DOI Creative Commons
Fangfang Gou, Jun Liu,

Chunwen Xiao

et al.

Diagnostics, Journal Year: 2024, Volume and Issue: 14(14), P. 1472 - 1472

Published: July 9, 2024

With the improvement of economic conditions and increase in living standards, people's attention regard to health is also continuously increasing. They are beginning place their hopes on machines, expecting artificial intelligence (AI) provide a more humanized medical environment personalized services, thus greatly expanding supply bridging gap between resource demand. development IoT technology, arrival 5G 6G communication era, enhancement computing capabilities particular, application AI-assisted healthcare have been further promoted. Currently, research field assistance deepening expanding. AI holds immense value has many potential applications institutions, patients, professionals. It ability enhance efficiency, reduce costs, improve quality intelligent service experience for professionals patients. This study elaborates history timelines field, types technologies informatics, opportunities challenges medicine. The combination profound impact human life, improving levels life changing lifestyles.

Language: Английский

Citations

28
Semi-supervised recognition for artificial intelligence assisted pathology image diagnosis DOI Creative Commons
Pan Yao, Fangfang Gou,

Chunwen Xiao

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Sept. 20, 2024

The analysis and interpretation of cytopathological images are crucial in modern medical diagnostics. However, manually locating identifying relevant cells from the vast amount image data can be a daunting task. This challenge is particularly pronounced developing countries where there may shortage expertise to handle such tasks. acquiring large amounts high-quality labelled remains, many researchers have begun use semi-supervised learning methods learn unlabeled data. Although current models partially solve issue limited data, they inefficient exploiting samples. To address this, we introduce new AI-assisted scheme, Reliable-Unlabeled Semi-Supervised Segmentation (RU3S) model. model integrates ResUNet-SE-ASPP-Attention (RSAA) model, which includes Squeeze-and-Excitation (SE) network, Atrous Spatial Pyramid Pooling (ASPP) structure, Attention module, ResUNet architecture. Our leverages effectively, improving accuracy significantly. A novel confidence filtering strategy introduced make better samples, addressing scarcity Experimental results show 2.0% improvement mIoU over state-of-the-art segmentation ST, demonstrating our approach's effectiveness solving this problem.

Language: Английский

Citations

5
Pathological Image Segmentation Method Based on Multiscale and Dual Attention DOI Creative Commons
Jia Wu, Yuxia Niu, Ziqiang Ling

et al.

International Journal of Intelligent Systems, Journal Year: 2024, Volume and Issue: 2024(1)

Published: Jan. 1, 2024

Medical images play a significant part in biomedical diagnosis, but they have feature. The medical images, influenced by factors such as imaging equipment limitations, local volume effect, and others, inevitably exhibit issues like noise, blurred edges, inconsistent signal strength. These imperfections pose challenges create obstacles for doctors during their diagnostic processes. To address these issues, we present pathology image segmentation technique based on the multiscale dual attention mechanism (MSDAUnet), which consists of three primary components. Firstly, an denoising enhancement module is constructed using dynamic residual color histogram to remove noise improve clarity. Then, propose (DAM), extracts messages from both channel spatial dimensions, obtains key features, makes edge lesion area clearer. Finally, capturing information process addresses issue uneven strength certain extent. Each combined automatic pathological segmentation. Compared with traditional typical U‐Net model, MSDAUnet has better performance. On dataset provided Research Center Artificial Intelligence Monash University, IOU index high 72.7%, nearly 7% higher than that U‐Net, DSC 84.9%, also about U‐Net.

Language: Английский

Citations

3
Diffpvt:information filtering based diffusion model with PVT for medical image segmentation DOI Creative Commons

Chengming Wang,

Genji Yuan, Mengjun Li

et al.

International Journal of Machine Learning and Cybernetics, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 4, 2025

Language: Английский

Citations

0
TransRNetFuse: a highly accurate and precise boundary FCN-transformer feature integration for medical image segmentation DOI Creative Commons

Baotian Li,

Jing Zhou,

Fangfang Gou

et al.

Complex & Intelligent Systems, Journal Year: 2025, Volume and Issue: 11(5)

Published: March 17, 2025

Language: Английский

Citations

0
Enhanced boundary-directed lightweight approach for digital pathological image analysis in critical oncological diagnostics DOI

Ou Luo,

Jing Zhou,

Fangfang Gou

et al.

Journal of X-Ray Science and Technology, Journal Year: 2025, Volume and Issue: unknown

Published: April 22, 2025

Background Pathological images play a crucial role in the diagnosis of critically ill cancer patients. Since patients often seek medical assistance when their condition is severe, doctors face urgent challenge completing accurate diagnoses and developing surgical plans within limited timeframe. The complexity diversity pathological require significant investment time from specialized physicians for processing analysis, which can lead to missing optimal treatment window. Purpose Current decision support systems are challenged by high computational deep learning models, demand extensive data training, making it difficult meet real-time needs emergency diagnostics. Method This study addresses issue malignant bone tumors such as osteosarcoma proposing Lightened Boundary-enhanced Digital Image Recognition Strategy (LB-DPRS). strategy optimizes self-attention mechanism Transformer model innovatively implements boundary segmentation enhancement strategy, thereby improving recognition accuracy tissue backgrounds nuclear boundaries. Additionally, this research introduces row-column attention methods sparsify matrix, reducing burden enhancing speed. Furthermore, proposed complementary further assists convolutional layers fully extracting detailed features . Results DSC value LB-DPRS reached 0.862, IOU 0.749, params was only 10.97 M. Conclusion Experimental results demonstrate that significantly improves efficiency while maintaining prediction interpretability, providing powerful efficient osteosarcoma.

Language: Английский

Citations

0
Intelligent cell images segmentation system: based on SDN and moving transformer DOI Creative Commons
Jia Wu, Pan Yao, Qing Ye

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Oct. 22, 2024

Diagnosing diseases heavily relies on cell pathology images, but the extensive data in each manual identification of relevant cells labor-intensive, especially regions with a scarcity qualified healthcare professionals. This study aims to develop an intelligent system enhance diagnostic accuracy cytopathology images by addressing image noise and segmentation issues, thereby improving efficiency medical professionals disease diagnosis. We introduced innovative combining self-supervised algorithm, SDN, for denoising enhancement using UPerMVit model. The model's novel attention mechanisms modular architecture provide higher lower computational complexity than traditional methods. proposed effectively reduces accurately segments annotated highlighting cellular structures staff. enhances aids accurate pathological cells. Our offers reliable tool professionals, cytopathologic analysis. It provides significant technical support lacking adequate expertise.

Language: Английский

Citations

1
FASNet: Feature Alignment-based method with digital pathology images in assisted diagnosis medical system DOI Creative Commons

Keke He,

Jun Zhu, Limiao Li

et al.

Heliyon, Journal Year: 2024, Volume and Issue: 10(22), P. e40350 - e40350

Published: Nov. 1, 2024

Many important information in medical research and clinical diagnosis are obtained from images. Among them, digital pathology images can provide detailed tissue structure cellular information, which has become the gold standard for tumor diagnosis. With development of neural networks, computer-aided presents identification results various cell nuclei to doctors, facilitates cancerous regions. However, deep learning models require a large amount annotated data. Pathology expensive difficult obtain, insufficient annotation data easily lead biased results. In addition, when current evaluated on an unknown target domain, there errors predicted boundaries. Based this, this study proposes feature alignment-based detail recognition strategy image segmentation (FASNet). It consists preprocessing model network (UNW). The UNW performs instance normalization categorical whitening by inserting semantics-aware modules into encoder decoder, achieves compactness features same class separation different classes. FASNet method identify more efficiently, thus differentiate between classes tissues effectively. experimental show that Dice Similarity Coefficient (DSC) value 0.844. good performance even faced with test does not match distribution training Code: https://github.com/zlf010928/FASNet.git.

Language: Английский

Citations

1