人工智能影像与多组学推动克罗恩病术后复发个体化管理
术后复发影响高达70%的克罗恩病患者,现有监测手段准确性不足且缺乏标准化评分与监测间隔。综述总结高分辨率内镜、肠道超声及多组学进展,探讨人工智能整合多模态数据实现风险分层、早期检测与个体化治疗的前景。
要点
- 术后第一年内克罗恩病术后复发率可达70%。
- 标准回结肠镜与粪钙卫蛋白监测准确性有限,缺乏验证评分与标准间隔。
- AI整合临床、内镜、影像与组学数据可支持个体化风险分层。
核心结论
- 人工智能整合内镜、影像与多组学数据有望推动克罗恩病术后复发的精准医学管理。
研究设计与样本
- 本篇为综述,总结克罗恩病术后复发的现有认识及新兴诊断与监测技术。
主要发现
- 术后复发影响高达70%的克罗恩病患者,多因素发病机制增加预防难度。
- 标准回结肠镜与粪钙卫蛋白准确性欠佳,缺乏验证评分系统与标准化监测间隔。
- 共聚焦激光显微内镜与内镜细胞镜可实时在体评估吻合口微结构。
- 肠道超声与横断面影像通过透壁评估重塑无创监测。
- 基因组学、转录组学、蛋白质组学、代谢组学与宏基因组学等发现与术后复发相关的新通路。
临床怎么用
- 上述创新需前瞻性验证、方法学标准化并整合入临床工作流程后方可常规应用。
局限性
- 新兴技术尚需前瞻性验证与方法学标准化,尚未整合入常规临床流程。
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
Postoperative recurrence (POR) is a major challenge in the long-term management of Crohn's disease (CD), affecting up to 70% of patients within the first year after surgical resection. The multifactorial pathogenesis of POR complicates prevention, while evolving surgical techniques and different anastomotic configurations further hinder accurate prediction and monitoring.Current surveillance strategies, including standard ileocolonoscopy and faecal calprotectin, remain limited by suboptimal accuracy, the absence of validated scoring systems and the lack of standardised monitoring intervals. Recent advances in high-resolution endoscopic imaging, such as confocal laser endomicroscopy and endocytoscopy, enable real-time, in vivo microstructural assessment of the anastomosis, offering opportunities for earlier and more precise detection of recurrence. In parallel, developments in intestinal ultrasound and cross-sectional imaging are reshaping non-invasive monitoring by providing transmural evaluation. Beyond imaging, multiomics approaches, spanning genomics, transcriptomics, proteomics, metabolomics and metagenomics, are uncovering novel biological pathways linked to POR, providing new mechanistic insights.Artificial intelligence (AI) has the potential to integrate clinical, endoscopic, imaging and omics data into predictive multimodal models for POR, supporting individualised risk stratification, early detection and personalised treatment strategies. While promising, these innovations require prospective validation, methodological standardisation and integration into clinical workflows before translation into routine practice.This review summarises the current understanding of POR, highlights emerging diagnostic and monitoring technologies and explores how AI-enabled endoscopy and multi-omics approaches may transform future management, paving the way towards precision medicine for POR in CD.
引用
- Iacucci M, Zammarchi I, Pugliano CL, et al. Shaping the future of postoperative recurrence in Crohn's disease: personalised approaches with AI-enabled imaging and multi-omics. Gut. 2026. PMID: 41592952. doi:10.1136/gutjnl-2025-337171
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。