Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy
Abstract This study was conducted to develop and validate a novel deep reinforcement learning (DRL) algorithm incorporating the segment anything model (SAM) to enhance the accuracy of automatic contouring organs at risk during radiotherapy for cervical cancer patients. CT images were collected from...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-91362-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850026106583777280 |
|---|---|
| author | Li Yucheng Qiu Lingyun Shao Kainan Jia Yongshi Zhan Wenming Ding Jieni Chen Weijun |
| author_facet | Li Yucheng Qiu Lingyun Shao Kainan Jia Yongshi Zhan Wenming Ding Jieni Chen Weijun |
| author_sort | Li Yucheng |
| collection | DOAJ |
| description | Abstract This study was conducted to develop and validate a novel deep reinforcement learning (DRL) algorithm incorporating the segment anything model (SAM) to enhance the accuracy of automatic contouring organs at risk during radiotherapy for cervical cancer patients. CT images were collected from 150 cervical cancer patients treated at our hospital between 2021 and 2023. Among these images, 122 CT images were used as a training set for the algorithm training of the DRL model based on the SAM model, and 28 CT images were used for the test set. The model’s performance was evaluated by comparing its segmentation results with the ground truth (manual contouring) obtained through manual contouring by expert clinicians. The test results were compared with the contouring results of commercial automatic contouring software based on the deep learning (DL) algorithm model. The Dice similarity coefficient (DSC), 95th percentile Hausdorff distance, average symmetric surface distance (ASSD), and relative absolute volume difference (RAVD) were used to quantitatively assess the contouring accuracy from different perspectives, enabling the contouring results to be comprehensively and objectively evaluated. The DRL model outperformed the DL model across all evaluated metrics. DRL achieved higher median DSC values, such as 0.97 versus 0.96 for the left kidney (P < 0.001), and demonstrated better boundary accuracy with lower HD95 values, e.g., 14.30 mm versus 17.24 mm for the rectum (P < 0.001). Moreover, DRL exhibited superior spatial agreement (median ASSD: 1.55 mm vs. 1.80 mm for the rectum, P < 0.001) and volume prediction accuracy (median RAVD: 10.25 vs. 10.64 for the duodenum, P < 0.001). These findings indicate that integrating SAM with RL (reinforcement learning) enhances segmentation accuracy and consistency compared to conventional DL methods. The proposed approach introduces a novel training strategy that improves performance without increasing model complexity, demonstrating its potential applicability in clinical practice. |
| format | Article |
| id | doaj-art-8fd4f1dd341c4293ac5a7a3f11c72e50 |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-8fd4f1dd341c4293ac5a7a3f11c72e502025-08-20T03:00:39ZengNature PortfolioScientific Reports2045-23222025-02-0115111110.1038/s41598-025-91362-9Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapyLi Yucheng0Qiu Lingyun1Shao Kainan2Jia Yongshi3Zhan Wenming4Ding Jieni5Chen Weijun6Cancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeCancer Center, Department of Radiation Oncology, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical CollegeAbstract This study was conducted to develop and validate a novel deep reinforcement learning (DRL) algorithm incorporating the segment anything model (SAM) to enhance the accuracy of automatic contouring organs at risk during radiotherapy for cervical cancer patients. CT images were collected from 150 cervical cancer patients treated at our hospital between 2021 and 2023. Among these images, 122 CT images were used as a training set for the algorithm training of the DRL model based on the SAM model, and 28 CT images were used for the test set. The model’s performance was evaluated by comparing its segmentation results with the ground truth (manual contouring) obtained through manual contouring by expert clinicians. The test results were compared with the contouring results of commercial automatic contouring software based on the deep learning (DL) algorithm model. The Dice similarity coefficient (DSC), 95th percentile Hausdorff distance, average symmetric surface distance (ASSD), and relative absolute volume difference (RAVD) were used to quantitatively assess the contouring accuracy from different perspectives, enabling the contouring results to be comprehensively and objectively evaluated. The DRL model outperformed the DL model across all evaluated metrics. DRL achieved higher median DSC values, such as 0.97 versus 0.96 for the left kidney (P < 0.001), and demonstrated better boundary accuracy with lower HD95 values, e.g., 14.30 mm versus 17.24 mm for the rectum (P < 0.001). Moreover, DRL exhibited superior spatial agreement (median ASSD: 1.55 mm vs. 1.80 mm for the rectum, P < 0.001) and volume prediction accuracy (median RAVD: 10.25 vs. 10.64 for the duodenum, P < 0.001). These findings indicate that integrating SAM with RL (reinforcement learning) enhances segmentation accuracy and consistency compared to conventional DL methods. The proposed approach introduces a novel training strategy that improves performance without increasing model complexity, demonstrating its potential applicability in clinical practice.https://doi.org/10.1038/s41598-025-91362-9Deep reinforcement learningSegment anything modelOrgans at riskCervical cancerDeep learning |
| spellingShingle | Li Yucheng Qiu Lingyun Shao Kainan Jia Yongshi Zhan Wenming Ding Jieni Chen Weijun Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy Scientific Reports Deep reinforcement learning Segment anything model Organs at risk Cervical cancer Deep learning |
| title | Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy |
| title_full | Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy |
| title_fullStr | Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy |
| title_full_unstemmed | Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy |
| title_short | Development and validation of a deep reinforcement learning algorithm for auto-delineation of organs at risk in cervical cancer radiotherapy |
| title_sort | development and validation of a deep reinforcement learning algorithm for auto delineation of organs at risk in cervical cancer radiotherapy |
| topic | Deep reinforcement learning Segment anything model Organs at risk Cervical cancer Deep learning |
| url | https://doi.org/10.1038/s41598-025-91362-9 |
| work_keys_str_mv | AT liyucheng developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT qiulingyun developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT shaokainan developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT jiayongshi developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT zhanwenming developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT dingjieni developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy AT chenweijun developmentandvalidationofadeepreinforcementlearningalgorithmforautodelineationoforgansatriskincervicalcancerradiotherapy |