基于三维卷积神经网络的尿道前列腺段自动分割AI模型开发
本研究纳入939例T2加权MRI扫描(训练/验证/测试/排除:657/141/140/1例),基于3D-UNet开发尿道前列腺段全自动分割模型,总体Dice评分0.61、中心线距离2.56 mm;模型在对称性良性前列腺增生腺体上表现优于非对称性腺体,且前列腺尖部表现最佳、基底部最差。
要点
- 模型总体分割性能Dice 0.61、中心线距离2.56 mm。
- 对称性良性前列腺增生腺体Dice 0.64 vs 非对称性0.51,模型在对称腺体上更优。
- 分割性能尖部最高、基底部最低,非对称腺体基底部 Dice 仅 0.39。
核心结论
- 研究开发了尿道前列腺段全自动分割AI工具,在具有对称性良性前列腺增生特征的腺体中表现最佳。
- 该系统可能用于辅助可接受全腺体放疗或消融局灶治疗患者的治疗计划。
研究设计与样本
- 研究纳入939例患者的T2加权MRI扫描,含内部数据集和公开PROSTATE-x数据集,训练/验证/测试/排除分别为657/141/140/1例。
- 以泌尿生殖放射科专家勾画的尿道轮廓作为参考标准。
- AI模型基于MONAI框架和3D-UNet开发。
- 性能采用基于体积的Dice评分和基于层面的预测与真实中心之间的中心线距离(CLD)评估,并进行系统失败分析,组间比较采用Wilcoxon秩和检验。
诊断标准与关键数值
- 总体器官校正Dice评分为0.61,总体CLD为2.56 mm。
- 对称性良性前列腺增生(n=117)vs 非对称性(n=23):Dice 0.64 vs 0.51(p < 0.05),平均CLD 2.3 mm vs 3.8 mm(p < 0.05)。
- Dice位置依赖性:对称性腺体尖部/中部/基底部分别为0.69 vs 0.67 vs 0.54(p < 0.05);非对称性腺体分别为0.68 vs 0.52 vs 0.39(p < 0.05)。
- CLD位置依赖性:对称性腺体尖部/中部/基底部分别为1.43 mm vs 2.15 mm vs 3.28 mm(p < 0.05);非对称性腺体分别为1.83 mm vs 3.1 mm vs 6.24 mm(p < 0.05)。
临床应用与阅片要点
- 模型在对称性良性前列腺增生腺体上分割表现优于非对称性腺体,阅片时需注意非对称腺体的分割可靠性可能较低。
- 前列腺尖部分割表现最佳、基底部最差,基底部尤其是非对称腺体基底部的分割结果需重点核对。
局限性
- 原文未报告局限性。
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)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVE: The combined use of prostate cancer radiotherapy and MRI planning is increasingly being used in the treatment of clinically significant prostate cancers. The radiotherapy dosage quantity is limited by toxicity in organs with de-novo genitourinary toxicity occurrence remaining unperturbed. Estimation of the urethral radiation dose via anatomical contouring may improve our understanding of genitourinary toxicity and its related symptoms. Yet, urethral delineation remains an expert-dependent and time-consuming procedure. In this study, we aim to develop a fully automated segmentation tool for the prostatic urethra.
MATERIALS AND METHODS: This study incorporated 939 patients' T2-weighted MRI scans (train/validation/test/excluded: 657/141/140/1 patients), including in-house and public PROSTATE-x datasets, and their corresponding ground truth urethral contours from an expert genitourinary radiologist. The AI model was developed using MONAI framework and was based on a 3D-UNet. AI model performance was determined by Dice score (volume-based) and the Centerline Distance (CLD) between the prediction and ground truth centers (slice-based). All predictions were compared to ground truth in a systematic failure analysis to elucidate the model's strengths and weaknesses. The Wilcoxon-rank sum test was used for pair-wise comparison of group differences.
RESULTS: The overall organ-adjusted Dice score for this model was 0.61 and overall CLD was 2.56 mm. When comparing prostates with symmetrical (n = 117) and asymmetrical (n = 23) benign prostate hyperplasia (BPH), the AI model performed better on symmetrical prostates compared to asymmetrical in both Dice score (0.64 vs. 0.51 respectively, p < 0.05) and mean CLD (2.3 mm vs. 3.8 mm respectively, p < 0.05). When calculating location-specific performance, the performance was highest at the apex and lowest at the base location of the prostate for Dice and CLD. Dice location dependence: symmetrical (Apex, Mid, Base: 0.69 vs. 0.67 vs. 0.54 respectively, p < 0.05) and asymmetrical (Apex, Mid, Base: 0.68 vs. 0.52 vs. 0.39 respectively, p < 0.05). CLD location dependence: symmetrical (Apex, Mid, Base: 1.43 mm vs. 2.15 mm vs. 3.28 mm, p < 0.05) and asymmetrical (Apex, Mid, Base: 1.83 mm vs. 3.1 mm vs. 6.24 mm, p < 0.05).
CONCLUSION: We developed a fully automated prostatic urethra segmentation AI tool yielding its best performance in prostate glands with symmetric BPH features. This system can potentially be used to assist treatment planning in patients who can undergo whole gland radiation therapy or ablative focal therapy.
引用
- Belue, Harmon, Patel, et al. Development of a 3D CNN-based AI Model for Automated Segmentation of the Prostatic Urethra. Acad Radiol. 2022. PMID: 35183438. doi:10.1016/j.acra.2022.01.009
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。