Machine Learning Model Predicts Postoperative Outcomes in Chronic Rhinosinusitis With Nasal Polyps

IF 1.7 4区医学 Q2 OTORHINOLARYNGOLOGY

Clinical Otolaryngology Pub Date : 2024-08-07 DOI:10.1111/coa.14208

Anda Gata, Lajos Raduly, Liviuța Budișan, Adél Bajcsi, Teodora-Maria Ursu, Camelia Chira, Laura Dioșan, Ioana Berindan-Neagoe, Silviu Albu

{"title":"Machine Learning Model Predicts Postoperative Outcomes in Chronic Rhinosinusitis With Nasal Polyps","authors":"Anda Gata, Lajos Raduly, Liviuța Budișan, Adél Bajcsi, Teodora-Maria Ursu, Camelia Chira, Laura Dioșan, Ioana Berindan-Neagoe, Silviu Albu","doi":"10.1111/coa.14208","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Objective</h3>\n \n <p>Evaluating the possibility of predicting chronic rhinosinusitis with nasal polyps (CRSwNP) disease course using Artificial Intelligence.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>We prospectively included patients undergoing first endoscopic sinus surgery (ESS) for nasal polyposis. Preoperative (demographic data, blood eosinophiles, endoscopy, Lund-Mackay, SNOT-22 and depression PHQ scores) and follow-up data was standardly collected. Outcome measures included SNOT-22, PHQ-9 and endoscopy perioperative sinus endoscopy (POSE) scores and two different microRNAs (miR-125b, miR-203a-3p) from polyp tissue. Based on POSE score, three labels were created (controlled: 0–7; partial control: 8–15; or relapse: 16–32). Patients were divided into train and test groups and using Random Forest, we developed algorithms for predicting ESS related outcomes.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>Based on data collected from 85 patients, the proposed Machine Learning-approach predicted whether the patient would present control, partial control or relapse of nasal polyposis at 18 months following ESS. The algorithm predicted ESS outcomes with an accuracy between 69.23% (for non-invasive input parameters) and 84.62% (when microRNAs were also included). Additionally, miR-125b significantly improved the algorithm's accuracy and ranked as one of the most important algorithm variables.</p>\n </section>\n \n <section>\n \n <h3> Conclusion</h3>\n \n <p>We propose a Machine Learning algorithm which could change the prediction of disease course in CRSwNP.</p>\n </section>\n </div>","PeriodicalId":10431,"journal":{"name":"Clinical Otolaryngology","volume":"49 6","pages":"776-784"},"PeriodicalIF":1.7000,"publicationDate":"2024-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/coa.14208","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Clinical Otolaryngology","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/coa.14208","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"OTORHINOLARYNGOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Objective

Evaluating the possibility of predicting chronic rhinosinusitis with nasal polyps (CRSwNP) disease course using Artificial Intelligence.

Methods

We prospectively included patients undergoing first endoscopic sinus surgery (ESS) for nasal polyposis. Preoperative (demographic data, blood eosinophiles, endoscopy, Lund-Mackay, SNOT-22 and depression PHQ scores) and follow-up data was standardly collected. Outcome measures included SNOT-22, PHQ-9 and endoscopy perioperative sinus endoscopy (POSE) scores and two different microRNAs (miR-125b, miR-203a-3p) from polyp tissue. Based on POSE score, three labels were created (controlled: 0–7; partial control: 8–15; or relapse: 16–32). Patients were divided into train and test groups and using Random Forest, we developed algorithms for predicting ESS related outcomes.

Results

Based on data collected from 85 patients, the proposed Machine Learning-approach predicted whether the patient would present control, partial control or relapse of nasal polyposis at 18 months following ESS. The algorithm predicted ESS outcomes with an accuracy between 69.23% (for non-invasive input parameters) and 84.62% (when microRNAs were also included). Additionally, miR-125b significantly improved the algorithm's accuracy and ranked as one of the most important algorithm variables.

Conclusion

We propose a Machine Learning algorithm which could change the prediction of disease course in CRSwNP.

Abstract Image

查看原文本刊更多论文

机器学习模型预测伴有鼻息肉的慢性鼻窦炎患者的术后效果

目的：评估利用人工智能预测慢性鼻息肉鼻炎（CRSwNP）病程的可能性：评估利用人工智能预测慢性鼻炎伴鼻息肉（CRSwNP）病程的可能性：我们前瞻性地纳入了因鼻息肉病首次接受内窥镜鼻窦手术（ESS）的患者。术前（人口统计学数据、血液嗜酸性粒细胞、内窥镜检查、Lund-Mackay、SNOT-22 和抑郁 PHQ 评分）和随访数据均按标准收集。结果测量包括 SNOT-22、PHQ-9 和内窥镜围手术期鼻窦内窥镜检查（POSE）评分，以及息肉组织中两种不同的 microRNA（miR-125b、miR-203a-3p）。根据 POSE 评分创建了三个标签（控制：0-7；部分控制：8-15；或复发：16-32）。患者被分为训练组和测试组，我们使用随机森林（Random Forest）开发了预测ESS相关结果的算法：结果：根据从 85 名患者收集的数据，所提出的机器学习方法可预测患者在 ESS 治疗 18 个月后鼻息肉是否会得到控制、部分控制或复发。该算法预测ESS结果的准确率在69.23%（非侵入性输入参数）和84.62%（微RNA也包括在内）之间。此外，miR-125b 显著提高了算法的准确性，并被列为最重要的算法变量之一：我们提出了一种机器学习算法，它可以改变对 CRSwNP 病程的预测。

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

Clinical Otolaryngology 医学-耳鼻喉科学

CiteScore

4.00

自引率

4.80%

发文量

106

审稿时长

>12 weeks

期刊介绍： Clinical Otolaryngology is a bimonthly journal devoted to clinically-oriented research papers of the highest scientific standards dealing with: current otorhinolaryngological practice audiology, otology, balance, rhinology, larynx, voice and paediatric ORL head and neck oncology head and neck plastic and reconstructive surgery continuing medical education and ORL training The emphasis is on high quality new work in the clinical field and on fresh, original research. Each issue begins with an editorial expressing the personal opinions of an individual with a particular knowledge of a chosen subject. The main body of each issue is then devoted to original papers carrying important results for those working in the field. In addition, topical review articles are published discussing a particular subject in depth, including not only the opinions of the author but also any controversies surrounding the subject. • Negative/null results In order for research to advance, negative results, which often make a valuable contribution to the field, should be published. However, articles containing negative or null results are frequently not considered for publication or rejected by journals. We welcome papers of this kind, where appropriate and valid power calculations are included that give confidence that a negative result can be relied upon.