■肛肠疾病和盆底疾病在普通人群中很普遍。患者可能出现重叠症状,延迟诊断,降低生活质量。治疗医生遇到许多挑战归因于骨盆解剖的复杂性,诊断技术的局限性,缺乏可用资源。本文概述了人工智能(AI)在解决良性肛门直肠疾病和盆底疾病管理困难方面的现状。
■根据系统评价和荟萃分析指南的首选报告项目进行系统文献综述。我们搜索了PubMed数据库,以确定2000年1月至2023年8月发表的所有潜在相关研究。搜索查询使用以下术语构建:AI,机器学习,深度学习,良性肛肠疾病,盆底疾病,大便失禁,阻塞性排便,肛瘘,直肠脱垂,和肛门直肠测压.排除了恶性肛门直肠文章和摘要。对所选文章的数据进行分析。
■共发现139篇文章,其中15个符合我们的纳入和排除标准。最常见的AI模块是卷积神经网络。研究人员能够开发AI模块来优化骨盆的成像研究,瘘管,和脓肿解剖学,促进肛门直肠测压解释,并改进了高清肛门镜的使用。没有一个模块在外部队列中验证。
■AI有可能加强盆底和良性肛肠疾病的管理。正在进行的研究需要使用多学科方法以及医生和人工智能程序员之间的合作来应对紧迫的挑战。
UNASSIGNED: Anorectal diseases and pelvic floor disorders are prevalent among the general population. Patients may present with overlapping symptoms, delaying diagnosis, and lowering quality of life. Treating physicians encounter numerous challenges attributed to the complex nature of pelvic anatomy, limitations of diagnostic techniques, and lack of available resources. This article is an overview of the current state of artificial intelligence (AI) in tackling the difficulties of managing benign anorectal disorders and pelvic floor disorders.
UNASSIGNED: A systematic literature review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We searched the PubMed database to identify all potentially relevant studies published from January 2000 to August 2023. Search queries were built using the following terms: AI, machine learning, deep learning, benign anorectal disease, pelvic floor disorder, fecal incontinence, obstructive defecation, anal fistula, rectal prolapse, and anorectal manometry. Malignant anorectal articles and abstracts were excluded. Data from selected articles were analyzed.
UNASSIGNED: 139 articles were found, 15 of which met our inclusion and exclusion criteria. The most common AI module was convolutional neural network. researchers were able to develop AI modules to optimize imaging studies for pelvis, fistula, and abscess anatomy, facilitated anorectal manometry interpretation, and improved high-definition anoscope use. None of the modules were validated in an external cohort.
UNASSIGNED: There is potential for AI to enhance the management of pelvic floor and benign anorectal diseases. Ongoing research necessitates the use of multidisciplinary approaches and collaboration between physicians and AI programmers to tackle pressing challenges.