Yucheng Li, Lingyun Qiu, Kainan Shao, Yongshi Jia, Wenming Zhan, Jieni Ding, Weijun Chen
Cancer Center, Department of Radiation Oncology, Zhejiang Provincial People's Hospital, Affiliated People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang, China.
Sci Rep. 2025 Feb 25;15(1):6800. doi: 10.1038/s41598-025-91362-9.
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.
本研究旨在开发并验证一种结合分割一切模型（SAM）的新型深度强化学习（DRL）算法，以提高宫颈癌患者放疗期间危及器官自动轮廓勾画的准确性。收集了2021年至2023年期间在我院接受治疗的150例宫颈癌患者的CT图像。在这些图像中，122张CT图像用作基于SAM模型的DRL模型算法训练的训练集，28张CT图像用作测试集。通过将模型的分割结果与专家临床医生手动勾画获得的真实情况（手动轮廓勾画）进行比较，来评估模型的性能。将测试结果与基于深度学习（DL）算法模型的商业自动轮廓勾画软件的轮廓勾画结果进行比较。使用骰子相似系数（DSC）、第95百分位数豪斯多夫距离、平均对称表面距离（ASSD）和相对绝对体积差（RAVD）从不同角度定量评估轮廓勾画准确性，从而能够对轮廓勾画结果进行全面、客观的评估。在所有评估指标上，DRL模型均优于DL模型。DRL获得了更高的中位数DSC值，例如左肾的DSC值为0.97，而DL模型为0.96（P < 0.001），并且在边界准确性方面表现更好，HD95值更低，例如直肠的HD95值为14.30毫米，而DL模型为17.24毫米（P < 0.001）。此外，DRL在空间一致性方面表现更优（直肠的中位数ASSD：1.55毫米对1.80毫米，P < 0.001），在体积预测准确性方面也更优（十二指肠的中位数RAVD：10.25对10.64，P < 0.001）。这些发现表明，与传统的DL方法相比，将SAM与强化学习（RL）相结合可提高分割准确性和一致性。所提出的方法引入了一种新颖的训练策略，在不增加模型复杂性的情况下提高了性能，证明了其在临床实践中的潜在适用性。
