鼻科学中的虚拟手术与计算流体力学：范围综述

文献信息

中文摘要

背景： 鼻塞是患者就诊耳鼻喉科最常见的原因之一，然而旨在改善鼻腔通畅性的手术结局仍不一致，传统客观指标与主观感知通畅度之间的相关性尚不确定。计算流体力学（CFD）可基于颅面部计算机断层扫描模拟鼻腔气流；结合重建和修改鼻窦解剖结构的软件，可在切除任何组织之前以数字方式重现计划手术并预测其功能影响。本范围综述梳理了基于CFD的虚拟手术在鼻科学中的应用方式、模拟的验证方法以及所报告的临床结局。方法：遵循PRISMA范围综述扩展版，于2025年10月在PubMed和Scopus检索2015年至2025年发表的研究。纳入标准限定为涉及CFD虚拟手术工具、以人类或人源模型为对象的英文研究。排除儿童（<18岁）、睡眠呼吸暂停和正颌研究，以及呼吸道其他部位的研究和环境或毒理学研究。在Mendeley中去除重复文献，两名作者独立筛选标题和摘要。数据由两名作者提取。本综述未注册。结果：共纳入41项研究，涵盖鼻中隔、下鼻甲和中鼻甲、窦口鼻道复合体、上颌窦和蝶窦、鼻中隔穿孔以及术后粘连。仅41项研究中的7项将虚拟结果与实际术后解剖、气流或症状进行比较，且多数方面一致。在若干小型研究中，更大范围的切除并未进一步改善气流，反而降低了黏膜热量和水分交换、壁面剪切应力和区域冷却，但最佳切除范围取决于解剖结构和所考虑的结局。这些参数被认为比单纯阻力更能反映主观通畅感；但很少在患者中验证。最近的工作将机器学习和伴随优化作为规划支持的研究工具。结论：基于CFD的虚拟手术是规划鼻科手术范围的有前景工具。其常规应用需要标准化方法、前瞻性验证以及与患者报告结局（NOSE、SNOT-22）的相关性；仍缺乏具有此类终点的大型研究。

英文摘要

Background: Nasal obstruction is one of the most common reasons for which patients present to an otolaryngologist, yet the outcomes of operations aimed at improving nasal patency remain inconsistent, and the correlation between conventional objective measures and perceived patency is uncertain. Computational fluid dynamics (CFD) allows nasal airflow to be simulated from craniofacial computed tomography; combined with software that reconstructs and modifies sinonasal anatomy, it permits a planned procedure to be reproduced digitally and its functional impact predicted before any tissue is removed. This scoping review maps how CFD-based virtual surgery has been used in rhinology, how the simulations were validated, and which clinical outcomes were reported. Methods: Following the PRISMA extension for scoping reviews, PubMed and Scopus were searched in October 2025 for studies published between 2015 and 2025. Inclusion criteria restricted papers to English-language studies of virtual surgery tools involving CFD in humans or human-derived models. Pediatric (<18 years), sleep apnea, and orthognathic studies were excluded, as were studies of other parts of the respiratory tract and environmental or toxicological work. Duplicates were removed in Mendeley, and two authors independently screened titles and abstracts. Data were charted by two authors. The review was not registered. Results: Forty-one studies were included, covering the nasal septum, the inferior and middle turbinates, the ostiomeatal complex, the maxillary and sphenoid sinuses, septal perforations, and postoperative synechiae. Virtual results were compared with actual postoperative anatomy, airflow, or symptoms in only 7 of the 41 studies and agreed with them in most respects. In several small studies, a greater extent of resection did not improve airflow further and reduced mucosal heat and moisture exchange, wall shear stress, and regional cooling, but the extent that appeared best depended on the anatomy and the outcome considered. These parameters have been proposed as better correlates of the subjective sense of patency than resistance alone; this has rarely been tested in patients. The most recent work applies machine learning and adjoint optimization as research tools for planning support. Conclusions: CFD-based virtual surgery is a promising tool for planning the extent of rhinological procedures. Its routine use will require standardized methods, prospective validation, and correlation with patient-reported outcomes (NOSE, SNOT-22); larger studies with such endpoints are still lacking.