DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation
Abstract Intracranial haemorrhage (ICH) is an urgent and potentially fatal medical condition caused by brain blood vessel rupture, leading to blood accumulation in the brain tissue. Due to the pressure and damage it causes to brain tissue, ICH results in severe neurological impairment or even death....
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2024-12-01
|
| Series: | IET Systems Biology |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/syb2.12103 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846110615354998784 |
|---|---|
| author | Haiyan Liu Yu Zeng Hao Li Fuxin Wang Jianjun Chang Huaping Guo Jian Zhang |
| author_facet | Haiyan Liu Yu Zeng Hao Li Fuxin Wang Jianjun Chang Huaping Guo Jian Zhang |
| author_sort | Haiyan Liu |
| collection | DOAJ |
| description | Abstract Intracranial haemorrhage (ICH) is an urgent and potentially fatal medical condition caused by brain blood vessel rupture, leading to blood accumulation in the brain tissue. Due to the pressure and damage it causes to brain tissue, ICH results in severe neurological impairment or even death. Recently, deep neural networks have been widely applied to enhance the speed and precision of ICH detection yet they are still challenged by small or subtle hemorrhages. The authors introduce DDANet, a novel haematoma segmentation model for brain CT images. Specifically, a dilated convolution pooling block is introduced in the intermediate layers of the encoder to enhance feature extraction capabilities of middle layers. Additionally, the authors incorporate a self‐attention mechanism to capture global semantic information of high‐level features to guide the extraction and processing of low‐level features, thereby enhancing the model's understanding of the overall structure while maintaining details. DDANet also integrates residual networks, channel attention, and spatial attention mechanisms for joint optimisation, effectively mitigating the severe class imbalance problem and aiding the training process. Experiments show that DDANet outperforms current methods, achieving the Dice coefficient, Jaccard index, sensitivity, accuracy, and a specificity of 0.712, 0.601, 0.73, 0.997, and 0.998, respectively. The code is available at https://github.com/hpguo1982/DDANet. |
| format | Article |
| id | doaj-art-79c1face421c4f29aefa7a9cddce8b6c |
| institution | Kabale University |
| issn | 1751-8849 1751-8857 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Systems Biology |
| spelling | doaj-art-79c1face421c4f29aefa7a9cddce8b6c2024-12-23T18:41:56ZengWileyIET Systems Biology1751-88491751-88572024-12-0118628529710.1049/syb2.12103DDANet: A deep dilated attention network for intracerebral haemorrhage segmentationHaiyan Liu0Yu Zeng1Hao Li2Fuxin Wang3Jianjun Chang4Huaping Guo5Jian Zhang6Department of Neurology Xinyang Central Hospital Xinyang ChinaSchool of Computer and Information Techonology Xinyang Normal University Xinyang ChinaDepartment of Neurology Xinyang Central Hospital Xinyang ChinaDepartment of Neurology Xinyang Central Hospital Xinyang ChinaDepartment of Neurology Xinyang Central Hospital Xinyang ChinaSchool of Computer and Information Techonology Xinyang Normal University Xinyang ChinaSchool of Computer and Information Techonology Xinyang Normal University Xinyang ChinaAbstract Intracranial haemorrhage (ICH) is an urgent and potentially fatal medical condition caused by brain blood vessel rupture, leading to blood accumulation in the brain tissue. Due to the pressure and damage it causes to brain tissue, ICH results in severe neurological impairment or even death. Recently, deep neural networks have been widely applied to enhance the speed and precision of ICH detection yet they are still challenged by small or subtle hemorrhages. The authors introduce DDANet, a novel haematoma segmentation model for brain CT images. Specifically, a dilated convolution pooling block is introduced in the intermediate layers of the encoder to enhance feature extraction capabilities of middle layers. Additionally, the authors incorporate a self‐attention mechanism to capture global semantic information of high‐level features to guide the extraction and processing of low‐level features, thereby enhancing the model's understanding of the overall structure while maintaining details. DDANet also integrates residual networks, channel attention, and spatial attention mechanisms for joint optimisation, effectively mitigating the severe class imbalance problem and aiding the training process. Experiments show that DDANet outperforms current methods, achieving the Dice coefficient, Jaccard index, sensitivity, accuracy, and a specificity of 0.712, 0.601, 0.73, 0.997, and 0.998, respectively. The code is available at https://github.com/hpguo1982/DDANet.https://doi.org/10.1049/syb2.12103bioinformaticsimage segmentationlearning (artificial intelligence)medical image processingpatient diagnosis |
| spellingShingle | Haiyan Liu Yu Zeng Hao Li Fuxin Wang Jianjun Chang Huaping Guo Jian Zhang DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation IET Systems Biology bioinformatics image segmentation learning (artificial intelligence) medical image processing patient diagnosis |
| title | DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation |
| title_full | DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation |
| title_fullStr | DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation |
| title_full_unstemmed | DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation |
| title_short | DDANet: A deep dilated attention network for intracerebral haemorrhage segmentation |
| title_sort | ddanet a deep dilated attention network for intracerebral haemorrhage segmentation |
| topic | bioinformatics image segmentation learning (artificial intelligence) medical image processing patient diagnosis |
| url | https://doi.org/10.1049/syb2.12103 |
| work_keys_str_mv | AT haiyanliu ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT yuzeng ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT haoli ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT fuxinwang ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT jianjunchang ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT huapingguo ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation AT jianzhang ddanetadeepdilatedattentionnetworkforintracerebralhaemorrhagesegmentation |