科研动态PubMed 摘要待译

Artificial intelligence in image-guided cervical interventional radiotherapy: a systematic review.

PubMed 学科轮询 · 2026-10-15

完整摘要 · 中文译文

背景:影像引导介入放疗(IG-IRT)是局部晚期宫颈癌的标准治疗方案,但仍耗时且依赖操作者。人工智能(AI)已成为提高效率和可重复性的有前景的工具。 方法:按照PRISMA 2020指南进行系统综述,检索PubMed、Scopus和Web of Science中2010年至2026年发表的关于AI在宫颈IRT中影像应用的研究。纳入的研究包括评估分割(靶区和危及器官)、计划设计、重建、图像配准和施源器选择的原始研究。 结果:共纳入52项研究。基于深度学习的分割在不同解剖结构和不同研究中表现出不同的性能,在表现最佳的模型中,某些危及器官和靶区的Dice系数接近或超过0.90。个别研究报告了更短的处理时间、减少的勾画变异性、准确的施源器重建以及对施源器选择的潜在支持。然而,显著的方法学和临床异质性阻碍了AI方法之间的直接比较。 结论:现有研究表明,AI可能支持影像引导宫颈介入放疗中部分工作流程步骤的自动化,并辅助治疗相关决策。虽然若干应用已取得较高的技术性能,但大多数仍处于概念验证阶段,尚需稳健的外部验证和前瞻性 在常规临床采用之前,需要进行临床评估。

完整摘要 · 英文原文

BACKGROUND: Imaging-guided interventional radiotherapy (IG-IRT) is the standard of care for locally advanced cervical cancer but remains time-consuming and operator-dependent. Artificial intelligence (AI) has emerged as a promising tool to improve efficiency and reproducibility. METHODS: A systematic review was conducted according to PRISMA 2020 guidelines, searching PubMed, Scopus, and Web of Science for studies published between 2010 and 2026 investigating AI-based imaging applications in cervical IRT. Eligible studies included original research evaluating segmentation (targets and organ at risk), planning, reconstruction, image registration and applicator selection. RESULTS: Fifty-two studies were included. Deep learning-based segmentation showed variable performance across anatomical structures and studies, with Dice coefficients approaching or exceeding 0.90 for some organs at risk and target volumes in the best-performing models. Individual studies reported shorter processing times, reduced contouring variability, accurate applicator reconstruction, and potential support for applicator selection. However, the substantial methodological and clinical heterogeneity precluded direct comparisons among AI approaches. CONCLUSIONS: The available studies suggest that AI may support the automation of selected workflow steps and assist treatment-related decision-making in image-guided cervical interventional radiotherapy. While several applications have achieved high technical performance, most remain at the proof-of-concept stage, and robust external validation and prospective clinical evaluation are needed before routine clinical adoption.

摘要 待译

BACKGROUND: Imaging-guided interventional radiotherapy (IG-IRT) is the standard of care for locally advanced cervical cancer but remains time-consuming and operator-dependent. Artificial intelligence (AI) has emerged as a promising tool to improve efficiency and reproducibility. METHODS: A systematic review was conducted according to PRISMA 2020 guidelines, searching PubMed, Scopus, and Web of Science for studies published between 2010 and 2026 investigating AI-based imaging applications in cervical IRT.…

English
Artificial intelligence in image-guided cervical interventional radiotherapy: a systematic review.
内容仅供参考,不构成诊疗建议,不能替代执业医师的诊断。

← 返回元界