智能轮廓:深度学习驱动下基于4D CT最大和平均强度投影的非小细胞肺癌内大体肿瘤体积勾画
本研究回顾性纳入124例非小细胞肺癌患者,利用4D CT生成的最大和平均强度投影图像训练U-net、attention U-net和V-net模型自动勾画内大体肿瘤体积。attention U-net在平均强度投影上平均DSC为0.871±0.048,HD95为2.958±2.266 mm;在最大强度投影上平均DSC为0.852±0.053,HD95为3.209±2.136 mm。attention U-net与U-net结果相似,显著优于V-net。
要点
- attention U-net在AIP图像上平均DSC为0.871±0.048,HD95为2.958±2.266 mm。
- attention U-net在MIP图像上平均DSC为0.852±0.053,HD95为3.209±2.136 mm。
- attention U-net与U-net结果相似,显著优于V-net。
核心结论
- 深度学习模型可利用MIP和AIP图像自动勾画IGTV,简化勾画流程并提高肺癌放疗计划的准确性和一致性。
研究设计与样本
- 回顾性纳入124例NSCLC患者,按70%(n=87)和30%(n=37)分为训练和验证队列。
- 采集4D CT图像并生成相应的MIP和AIP图像,以4D CT上勾画的IGTV为金标准。
- 将MIP或AIP图像及对应IGTV输入U-net、attention U-net和V-net模型进行训练和验证。
主要发现
- attention U-net模型在AIP图像上平均DSC为0.871±0.048,平均HD95为2.958±2.266 mm。
- attention U-net模型在MIP图像上平均DSC为0.852±0.053,平均HD95为3.209±2.136 mm。
- attention U-net和U-net取得相似结果,显著优于V-net。
临床怎么用
- DL模型可利用MIP和AIP图像自动勾画IGTV,简化勾画流程。
- DL模型有助提高肺癌放疗计划的准确性和一致性,改善患者结局。
局限性
- 原文未述局限性
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: Delineating the internal gross tumor volume (IGTV) is crucial for the treatment of non-small cell lung cancer (NSCLC). Deep learning (DL) enables the automation of this process; however, current studies focus mainly on multiple phases of four-dimensional (4D) computed tomography (CT), which leads to indirect results. This study proposed a DL-based method for automatic IGTV delineation using maximum and average intensity projections (MIP and AIP, respectively) from 4D CT. METHODS: We retrospectively enrolled 124 patients with NSCLC and divided them into training (70%, n = 87) and validation (30%, n = 37) cohorts. Four-dimensional CT images were acquired, and the corresponding MIP and AIP images were generated. The IGTVs were contoured on 4D CT and used as the ground truth (GT). The MIP or AIP images, along with the corresponding IGTVs (IGTVMIP-manu and IGTVAIP-manu, respectively), were fed into the DL models for training and validation. We assessed the performance of three segmentation models-U-net, attention U-net, and V-net-using the Dice similarity coefficient (DSC) and the 95th percentile of the Hausdorff distance (HD95) as the primary metrics. RESULTS: The attention U-net model trained on AIP images presented a mean DSC of 0.871 ± 0.048 and mean HD95 of 2.958 ± 2.266 mm, whereas the model trained on MIP images achieved a mean DSC of 0.852 ± 0.053 and mean HD95 of 3.209 ± 2.136 mm. Among the models, attention U-net and U-net achieved similar results, considerably surpassing V-net. CONCLUSIONS: DL models can automate IGTV delineation using MIP and AIP images, streamline contouring, and enhance the accuracy and consistency of lung cancer radiotherapy planning to improve patient outcomes.
引用
- Huang Y, Luo M, Luo Z, et al. Smart contours: deep learning-driven internal gross tumor volume delineation in non-small cell lung cancer using 4D CT maximum and average intensity projections. Radiat Oncol. 2025. PMID: 40251610. doi:10.1186/s13014-025-02642-7
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。