肿瘤预防与治疗2025，Vol.38Issue(12)：1050-1059,10.DOI:10.3969/j.issn.1674-0904.2025.12.003
基于结直肠无蒂锯齿状病变白光内镜下特征诊断模型的构建及验证
Development and External Validation of a Diagnostic Model for White-Light Endoscopic Features of Colorectal Sessile Serrated Lesions
摘要
Abstract
Objective:To develop and validate a diagnostic model for sessile serrated lesions(SSL)based on their white-light endoscopic(WLE)characteristics.Meth-ods:We conducted a retrospective analysis of patients who underwent polypectomy for SSLs or hyperplastic polyps(HP)at Chengdu Third People's Hospital between November 2022 and April 2024.This included 416 patients(464 lesions)with pathologically confirmed SSL and 229 patients(247 lesions)with HP.SSL and HP lesions were separately and randomly divided into training and validation sets at a ratio of 7:3.In the training set,WLE features significantly associated with SSL were first screened using univariate logistic re-gression(P<0.05).Variables for inclusion were then determined by employing forward and backward stepwise selec-tion.Finally,a logistic regression-based diagnostic model was constructed and visualized as a nomogram.In the validation set,the model's performance was evaluated using the AUC,along with accuracy,precision,sensitivity,specificity,and the F1-score.Two methods were used for validation.First,discrimination and calibration were assessed by calculating the C-index and generating calibration curves via bootstrapping(1,000 repetitions).Second,clinical utility was evaluated u-sing decision curve analysis.Results:Lesion location,cloud-like surface,lesion border,lesion surface color,mucus,and lesion size were independent predictors of SSL(all P<0.05).The diagnostic model demonstrated an AUC of 0.804(95%CI:0.764～0.843)in the training set and 0.823(95%CI:0.764～0.882)in the validation set.In the train-ing set,the model achieved an accuracy,precision,sensitivity,specificity,and F1-score of 0.789,0.832,0.849,0.676,and 0.841,respectively.In the validation set,the corresponding values were 0.803,0.847,0.859,0.690,and 0.852.The AUC showed no significant difference between the training and validation sets(P>0.05),indicating consistent performance.The model showed good discrimination,with a C-index of 0.804 in the training set and 0.827 in the validation set.Calibration curves indicated good agreement between predicted and observed outcomes(P>0.05).Furthermore,decision curve analysis confirmed the model's clinical utility by demonstrating a significant net benefit across a threshold probability range of 0.1 to 0.9.Conclusion:The diagnostic model developed from the WLE features of SSL demonstrates high performance and can aid endoscopists in differentiating SSL from HP during white-light endoscopy.关键词
无蒂锯齿状病变/增生性息肉/结直肠癌/诊断模型/白光内镜/列线图Key words
Sessile serrated lesion/Hyperplastic polyp/Colorectal cancer/Diagnostic model/White light endoscopy/Nomogram分类
医药卫生引用本文复制引用
Ma Xiaomei,Shu Shunqing,Zhang Chen,Liu Liu,Sun Xiaobin..基于结直肠无蒂锯齿状病变白光内镜下特征诊断模型的构建及验证[J].肿瘤预防与治疗,2025,38(12):1050-1059,10.基金项目
This study was supported by grants from Health Commission of Chengdu(No.2025006)and the Third People's Hospital of Chengdu(No.2023-PI-04). 成都市卫健委医学科研课题(编号:2025006) （No.2025006）
成都市第三人民医院PI项目(编号:2023-PI-04) （编号:2023-PI-04）
