Detection of cervical cell based on multi-scale spatial information
Abstract Cervical cancer poses a significant health risk to women. Deep learning methods can assist pathologists in quickly screening images of suspected lesion cells, greatly improving the efficiency of cervical cancer screening and diagnosis. However, existing deep learning methods rely solely on...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-025-87165-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832585938556944384 |
|---|---|
| author | Gang Li Xinyu Fan Chuanyun Xu Pengfei Lv Ru Wang Zihan Ruan Zheng Zhou Yang Zhang |
| author_facet | Gang Li Xinyu Fan Chuanyun Xu Pengfei Lv Ru Wang Zihan Ruan Zheng Zhou Yang Zhang |
| author_sort | Gang Li |
| collection | DOAJ |
| description | Abstract Cervical cancer poses a significant health risk to women. Deep learning methods can assist pathologists in quickly screening images of suspected lesion cells, greatly improving the efficiency of cervical cancer screening and diagnosis. However, existing deep learning methods rely solely on single-scale features and local spatial information, failing to effectively capture the subtle morphological differences between abnormal and normal cervical cells. To tackle this problem effectively, we propose a cervical cell detection method that utilizes multi-scale spatial information. This approach efficiently captures and integrates spatial information at different scales. Firstly, we design the Multi-Scale Spatial Information Augmentation Module (MSA), which captures global spatial information by introducing a multi-scale spatial information extraction branch during the feature extraction stage. Secondly, the Channel Attention Enhanced Module (CAE) is introduced to achieve channel-level weighted processing, dynamically optimizing each output feature using channel weights to focus on critical features. We use Sparse R-CNN as the baseline and integrate MSA and CAE into it. Experiments on the CDetector dataset achieved an Average Precision (AP) of 65.3%, outperforming the state-of-the-art (SOTA) methods. |
| format | Article |
| id | doaj-art-ae1d7640ef044e739611da2c472c943e |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-ae1d7640ef044e739611da2c472c943e2025-01-26T12:24:20ZengNature PortfolioScientific Reports2045-23222025-01-0115111210.1038/s41598-025-87165-7Detection of cervical cell based on multi-scale spatial informationGang Li0Xinyu Fan1Chuanyun Xu2Pengfei Lv3Ru Wang4Zihan Ruan5Zheng Zhou6Yang Zhang7School of Artificial Intelligence, Chongqing University of TechnologySchool of Artificial Intelligence, Chongqing University of TechnologySchool of Computer and Information Science, Chongqing Normal UniversitySchool of Artificial Intelligence, Chongqing University of TechnologySchool of Artificial Intelligence, Chongqing University of TechnologySchool of Artificial Intelligence, Chongqing University of TechnologySchool of Artificial Intelligence, Chongqing University of TechnologySchool of Computer and Information Science, Chongqing Normal UniversityAbstract Cervical cancer poses a significant health risk to women. Deep learning methods can assist pathologists in quickly screening images of suspected lesion cells, greatly improving the efficiency of cervical cancer screening and diagnosis. However, existing deep learning methods rely solely on single-scale features and local spatial information, failing to effectively capture the subtle morphological differences between abnormal and normal cervical cells. To tackle this problem effectively, we propose a cervical cell detection method that utilizes multi-scale spatial information. This approach efficiently captures and integrates spatial information at different scales. Firstly, we design the Multi-Scale Spatial Information Augmentation Module (MSA), which captures global spatial information by introducing a multi-scale spatial information extraction branch during the feature extraction stage. Secondly, the Channel Attention Enhanced Module (CAE) is introduced to achieve channel-level weighted processing, dynamically optimizing each output feature using channel weights to focus on critical features. We use Sparse R-CNN as the baseline and integrate MSA and CAE into it. Experiments on the CDetector dataset achieved an Average Precision (AP) of 65.3%, outperforming the state-of-the-art (SOTA) methods.https://doi.org/10.1038/s41598-025-87165-7 |
| spellingShingle | Gang Li Xinyu Fan Chuanyun Xu Pengfei Lv Ru Wang Zihan Ruan Zheng Zhou Yang Zhang Detection of cervical cell based on multi-scale spatial information Scientific Reports |
| title | Detection of cervical cell based on multi-scale spatial information |
| title_full | Detection of cervical cell based on multi-scale spatial information |
| title_fullStr | Detection of cervical cell based on multi-scale spatial information |
| title_full_unstemmed | Detection of cervical cell based on multi-scale spatial information |
| title_short | Detection of cervical cell based on multi-scale spatial information |
| title_sort | detection of cervical cell based on multi scale spatial information |
| url | https://doi.org/10.1038/s41598-025-87165-7 |
| work_keys_str_mv | AT gangli detectionofcervicalcellbasedonmultiscalespatialinformation AT xinyufan detectionofcervicalcellbasedonmultiscalespatialinformation AT chuanyunxu detectionofcervicalcellbasedonmultiscalespatialinformation AT pengfeilv detectionofcervicalcellbasedonmultiscalespatialinformation AT ruwang detectionofcervicalcellbasedonmultiscalespatialinformation AT zihanruan detectionofcervicalcellbasedonmultiscalespatialinformation AT zhengzhou detectionofcervicalcellbasedonmultiscalespatialinformation AT yangzhang detectionofcervicalcellbasedonmultiscalespatialinformation |