Region-specific deep learning models for accurate segmentation of rectal structures on post-chemoradiation T2w MRI: a multi-institutional, multi-reader study.
Region-specific deep learning models for accurate segmentation of rectal structures on post-chemoradiation T2w MRI: a multi-institutional, multi-reader study.
复制标题
DOI：
10.3389/fmed.2023.1149056
复制
发表时间：
2023
影响因子：
3.9
通讯作者：
中科院分区：
文献类型：
--
作者：
对于局部晚期直肠癌，新辅助治疗后肿瘤范围和消退的体内放射学评估涉及磁共振成像（MRI）直肠结构的隐式视觉识别。此外，较新的基于图像的计算方法（例如放射组学）需要对直肠外壁、管腔和直肠周围脂肪等区域进行更详细和精确的注释。然而，这些区域的手动注释非常费力和耗时，并且由于组织边界被治疗相关的变化（例如纤维化、水肿）所掩盖，因此容易受到读者间差异的影响。这项研究展示了 U-Net 深度学习模型的应用，该模型是根据区域特定背景而独特开发的，可在治疗后 T2 加权 MRI 扫描中自动分割每个直肠外壁、管腔和直肠周围脂肪区域。在多机构评估中，发现特定区域的 U-Net（墙 Dice = 0.920，流明 Dice = 0.895）的性能与多个阅读器相当（墙阅读器间 Dice = 0.946，流明阅读器间 Dice = 0.873）。此外，与多类 U-Net 相比，特定区域的 U-Net 在分割壁、管腔和脂肪方面的 Dice 分数平均提高了 20%；即使在 T2 加权 MRI 扫描上进行测试时，图像质量较差，或者来自不同的平面，或者来自外部机构。因此，开发具有特定区域背景的深度学习分割模型可以在放化疗后 T2 加权 MRI 扫描上对多个直肠结构进行高度准确、详细的注释，这对于改善体内肿瘤范围的评估和为直肠癌构建准确的基于图像的分析工具至关重要。
For locally advanced rectal cancers, in vivo radiological evaluation of tumor extent and regression after neoadjuvant therapy involves implicit visual identification of rectal structures on magnetic resonance imaging (MRI). Additionally, newer image-based, computational approaches (e.g., radiomics) require more detailed and precise annotations of regions such as the outer rectal wall, lumen, and perirectal fat. Manual annotations of these regions, however, are highly laborious and time-consuming as well as subject to inter-reader variability due to tissue boundaries being obscured by treatment-related changes (e.g., fibrosis, edema). This study presents the application of U-Net deep learning models that have been uniquely developed with region-specific context to automatically segment each of the outer rectal wall, lumen, and perirectal fat regions on post-treatment, T2-weighted MRI scans. In multi-institutional evaluation, region-specific U-Nets (wall Dice = 0.920, lumen Dice = 0.895) were found to perform comparably to multiple readers (wall inter-reader Dice = 0.946, lumen inter-reader Dice = 0.873). Additionally, when compared to a multi-class U-Net, region-specific U-Nets yielded an average 20% improvement in Dice scores for segmenting each of the wall, lumen, and fat; even when tested on T2-weighted MRI scans that exhibited poorer image quality, or from a different plane, or were accrued from an external institution. Developing deep learning segmentation models with region-specific context may thus enable highly accurate, detailed annotations for multiple rectal structures on post-chemoradiation T2-weighted MRI scans, which is critical for improving evaluation of tumor extent in vivo and building accurate image-based analytic tools for rectal cancers.