Hou Zhen, Qin Lang, Gu Jiabing, Liu Zidong, Liu Juan, Zhang Yuan, Gao Shanbao, Zhu Jian, Li Shuangshuang
The Comprehensive Cancer Centre of Nanjing Drum Tower Hospital, The Affiliated Hospital of Nanjing University Medical School, Nanjing, Jiangsu, 210000, China.
The School of Electronic Science and Engineering, Nanjing University, Nanjing, Jiangsu, 210000, China.
Radiat Oncol. 2025 May 19;20(1):80. doi: 10.1186/s13014-025-02634-7.
Accurate pre-treatment dose prediction is essential for efficient radiotherapy planning. Although deep learning models have advanced automated dose distribution, comprehensive multi-tumor analyses remain scarce. This study assesses deep learning models for dose prediction across diverse tumor types, combining objective and subjective evaluation methods.
We included 622 patients with planning data across various tumor sites: nasopharyngeal carcinoma (n = 29), esophageal carcinoma (n = 82), left-sided breast carcinoma (n = 107), right-sided breast carcinoma (n = 95), cervical carcinoma treated with radical radiotherapy (n = 84), postoperative cervical carcinoma (n = 122), and rectal carcinoma (n = 103). Dose predictions were generated using U-Net, Flex-Net, and Highres-Net models, with data split into training (60%), validation (20%), and testing (20%) sets. Quantitative comparisons used normalized dose difference (NDD) and dose-volume histogram (DVH) metrics, and qualitative assessments by radiation oncologists were performed on the testing set.
Predicted and clinical doses correlated well, with NDD values under 3% for tumor targets in nasopharyngeal, breast, and postoperative cervical cancer. Qualitative assessments revealed that U-Net, Flex-Net, and Highres-Net achieved the highest accuracy in cervical radical, breast/rectal/postoperative cervical, and nasopharyngeal/esophageal cancers, respectively. Among the test cases (n = 123), 53.7% were deemed clinically acceptable and 32.5% required minor adjustments. The "Best Selection" approach, combining strengths of all three models, raised clinical acceptance to 62.6%.
This study demonstrates that automated dose prediction can provide a robust starting point for rapid plan generation. Leveraging model-specific strengths through the "Best Selection" approach enhances prediction accuracy and shows potential to improve clinical efficiency across multiple tumor types.
准确的预处理剂量预测对于高效的放射治疗计划至关重要。尽管深度学习模型推进了自动剂量分布，但全面的多肿瘤分析仍然很少。本研究结合客观和主观评估方法，评估深度学习模型在不同肿瘤类型中的剂量预测。
我们纳入了622例具有不同肿瘤部位计划数据的患者：鼻咽癌（n = 29）、食管癌（n = 82）、左侧乳腺癌（n = 107）、右侧乳腺癌（n = 95）、接受根治性放疗的宫颈癌（n = 84）、宫颈癌术后（n = 122）和直肠癌（n = 103）。使用U-Net、Flex-Net和Highres-Net模型生成剂量预测，数据分为训练集（60%）、验证集（20%）和测试集（20%）。定量比较使用归一化剂量差异（NDD）和剂量体积直方图（DVH）指标，并由放射肿瘤学家对测试集进行定性评估。
预测剂量与临床剂量相关性良好，鼻咽癌、乳腺癌和宫颈癌术后肿瘤靶区的NDD值低于3%。定性评估显示，U-Net、Flex-Net和Highres-Net分别在根治性宫颈癌、乳腺/直肠/宫颈癌术后和鼻咽癌/食管癌中达到了最高准确率。在测试病例（n = 123）中，53.7%被认为临床可接受，32.5%需要轻微调整。结合所有三种模型优势的“最佳选择”方法将临床接受率提高到了62.6%。
本研究表明，自动剂量预测可为快速计划生成提供一个可靠的起点。通过“最佳选择”方法利用模型特定优势可提高预测准确性，并显示出提高多种肿瘤类型临床效率的潜力。
