Machine Learning Prediction of Residual and Recurrent High-Grade CIN Post-LEEP

IF 2.5 4区医学 Q3 ONCOLOGY

Cancer Management and Research Pub Date : 2024-09-06 DOI:10.2147/cmar.s484057

Furui Zhai, Shanshan Mu, Yinghui Song, Min Zhang, Cui Zhang, Ze Lv

{"title":"Machine Learning Prediction of Residual and Recurrent High-Grade CIN Post-LEEP","authors":"Furui Zhai, Shanshan Mu, Yinghui Song, Min Zhang, Cui Zhang, Ze Lv","doi":"10.2147/cmar.s484057","DOIUrl":null,"url":null,"abstract":"Purpose: This study aims to develop a machine learning (ML) model to predict the risk of residual or recurrent high-grade cervical intraepithelial neoplasia (CIN) after loop electrosurgical excision procedure (LEEP), addressing a critical gap in personalized follow-up care. Methods: A retrospective analysis of 532 patients who underwent LEEP for high-grade CIN at Cangzhou Central Hospital (2016– 2020) was conducted. In the final analysis, 99 women (18.6%) were found to have residual or recurrent high-grade CIN (CIN2 or worse) within five years of follow-up. Four feature selection methods identified significant predictors of residual or recurrent CIN. Eight ML algorithms were evaluated using performance metrics such as AUROC, accuracy, sensitivity, specificity, PPV, NPV, F1 score, calibration curve, and decision curve analysis. Fivefold cross-validation optimized and validated the model, and SHAP analysis assessed feature importance. Results: The XGBoost algorithm demonstrated the highest predictive performance with the best AUROC. The optimized model included six key predictors: age, ThinPrep cytologic test (TCT) results, HPV classification, CIN severity, glandular involvement, and margin status. SHAP analysis identified CIN severity and margin status as the most influential predictors. An online prediction tool was developed for real-time risk assessment. Conclusion: This ML-based predictive model for post-LEEP high-grade CIN provides a significant advancement in gynecologic oncology, enhancing personalized patient care and facilitating early intervention and informed clinical decision-making. Keywords: cervical intraepithelial neoplasia, loop electrosurgical excision procedure, residual or recurrent, machine learning, predictive modeling ","PeriodicalId":9479,"journal":{"name":"Cancer Management and Research","volume":"44 1","pages":""},"PeriodicalIF":2.5000,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Cancer Management and Research","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.2147/cmar.s484057","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ONCOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Purpose: This study aims to develop a machine learning (ML) model to predict the risk of residual or recurrent high-grade cervical intraepithelial neoplasia (CIN) after loop electrosurgical excision procedure (LEEP), addressing a critical gap in personalized follow-up care.
Methods: A retrospective analysis of 532 patients who underwent LEEP for high-grade CIN at Cangzhou Central Hospital (2016– 2020) was conducted. In the final analysis, 99 women (18.6%) were found to have residual or recurrent high-grade CIN (CIN2 or worse) within five years of follow-up. Four feature selection methods identified significant predictors of residual or recurrent CIN. Eight ML algorithms were evaluated using performance metrics such as AUROC, accuracy, sensitivity, specificity, PPV, NPV, F1 score, calibration curve, and decision curve analysis. Fivefold cross-validation optimized and validated the model, and SHAP analysis assessed feature importance.
Results: The XGBoost algorithm demonstrated the highest predictive performance with the best AUROC. The optimized model included six key predictors: age, ThinPrep cytologic test (TCT) results, HPV classification, CIN severity, glandular involvement, and margin status. SHAP analysis identified CIN severity and margin status as the most influential predictors. An online prediction tool was developed for real-time risk assessment.
Conclusion: This ML-based predictive model for post-LEEP high-grade CIN provides a significant advancement in gynecologic oncology, enhancing personalized patient care and facilitating early intervention and informed clinical decision-making.

Keywords: cervical intraepithelial neoplasia, loop electrosurgical excision procedure, residual or recurrent, machine learning, predictive modeling

查看原文本刊更多论文

睡眠后残留和复发高级别 CIN 的机器学习预测

目的：本研究旨在开发一种机器学习（ML）模型，用于预测环状电切术（LEEP）后残留或复发高级别宫颈上皮内瘤变（CIN）的风险，解决个性化随访护理中的一个关键缺口：方法：对沧州市中心医院（2016-2020年）接受LEEP术治疗高级别CIN的532名患者进行回顾性分析。在最终分析中，发现99名女性（18.6%）在随访五年内有残留或复发高级别CIN（CIN2或更差）。四种特征选择方法确定了残留或复发 CIN 的重要预测因素。使用 AUROC、准确度、灵敏度、特异性、PPV、NPV、F1 分数、校准曲线和决策曲线分析等性能指标对八种 ML 算法进行了评估。五倍交叉验证对模型进行了优化和验证，SHAP分析评估了特征的重要性：结果：XGBoost 算法具有最高的预测性能和最佳的 AUROC。优化模型包括六个关键预测因素：年龄、ThinPrep 细胞学检测（TCT）结果、HPV 分类、CIN 严重程度、腺体受累和边缘状态。SHAP 分析确定 CIN 严重程度和边缘状态是最有影响力的预测因素。我们还开发了一个在线预测工具，用于实时风险评估：这个基于 ML 的 LEEP 术后高级别 CIN 预测模型为妇科肿瘤学提供了一个重要的进步，加强了对患者的个性化护理，促进了早期干预和知情的临床决策。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Cancer Management and Research Medicine-Oncology

CiteScore

7.40

自引率

0.00%

发文量

448

审稿时长

16 weeks

期刊介绍： Cancer Management and Research is an international, peer reviewed, open access journal focusing on cancer research and the optimal use of preventative and integrated treatment interventions to achieve improved outcomes, enhanced survival, and quality of life for cancer patients. Specific topics covered in the journal include: ◦Epidemiology, detection and screening ◦Cellular research and biomarkers ◦Identification of biotargets and agents with novel mechanisms of action ◦Optimal clinical use of existing anticancer agents, including combination therapies ◦Radiation and surgery ◦Palliative care ◦Patient adherence, quality of life, satisfaction The journal welcomes submitted papers covering original research, basic science, clinical & epidemiological studies, reviews & evaluations, guidelines, expert opinion and commentary, and case series that shed novel insights on a disease or disease subtype.