Machine Learning Applications in the Diagnosis and Prediction of Aortic Dissection: A Systematic Review and Meta-Analysis

IF 0.4 4区医学 Q3 CARDIAC & CARDIOVASCULAR SYSTEMS

Journal of Cardiac Surgery Pub Date : 2025-10-25 DOI:10.1155/jocs/3328006

Pooya Eini, Peyman Eini, Homa Serpoush, Mohammad Rezayee, Jason Tremblay

{"title":"Machine Learning Applications in the Diagnosis and Prediction of Aortic Dissection: A Systematic Review and Meta-Analysis","authors":"Pooya Eini, Peyman Eini, Homa Serpoush, Mohammad Rezayee, Jason Tremblay","doi":"10.1155/jocs/3328006","DOIUrl":null,"url":null,"abstract":"<div>\n \n <section>\n \n <h3> Background</h3>\n \n <p>Aortic dissection is a life-threatening condition requiring rapid and accurate diagnosis. Machine learning (ML) models have shown promise in enhancing diagnostic performance using imaging modalities, but their pooled efficacy remains unclear. This systematic review and meta-analysis evaluates the diagnostic accuracy of ML models for detecting aortic dissection.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>Following PRISMA guidelines, we searched PubMed, Scopus, Embase, Web of Science, and ProQuest up to March 10, 2025, identifying 775 articles. After removing duplicates, 358 articles were screened, 28 underwent full-text review, and 21 studies were included, with 18 providing sufficient metrics for meta-analysis. The MIDAS module in Stata was used to pool sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC) for the best-performing models, assessed via a bivariate mixed-effects model. Risk of bias was evaluated using PROBAST + AI.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>The 21 studies included 261,477 patients (mean age 58.74 years, 66.67% male), predominantly from China (11 studies). Imaging modalities included CT angiography (10 studies) and noncontrast/contrast-enhanced CT (five studies). The best models (18 studies, 7957 positive, 41,852 negative cases) achieved a pooled sensitivity of 0.92 (95% CI: 0.87–0.96), a specificity of 0.95 (95% CI: 0.91–0.98), and an AUROC of 0.98 (95% CI: 0.96–0.99). Deep learning models (DenseNet121, Attention U-Net) predominated. Moderate to substantial heterogeneity in both pooled sensitivity (<i>I</i><sup>2</sup> = 64.85%) and specificity (<i>I</i><sup>2</sup> = 62.87%), likely due to diverse ML algorithms and imaging protocols, was identified.</p>\n </section>\n \n <section>\n \n <h3> Conclusions</h3>\n \n <p>ML models show strong diagnostic performance for detecting aortic dissection, highlighting their promise for clinical use. However, the standardization of methodologies is essential to reduce variability and promote broader clinical implementation.</p>\n </section>\n </div>","PeriodicalId":15367,"journal":{"name":"Journal of Cardiac Surgery","volume":"2025 1","pages":""},"PeriodicalIF":0.4000,"publicationDate":"2025-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/jocs/3328006","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Cardiac Surgery","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1155/jocs/3328006","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"CARDIAC & CARDIOVASCULAR SYSTEMS","Score":null,"Total":0}

引用次数: 0

Abstract

Background

Aortic dissection is a life-threatening condition requiring rapid and accurate diagnosis. Machine learning (ML) models have shown promise in enhancing diagnostic performance using imaging modalities, but their pooled efficacy remains unclear. This systematic review and meta-analysis evaluates the diagnostic accuracy of ML models for detecting aortic dissection.

Methods

Following PRISMA guidelines, we searched PubMed, Scopus, Embase, Web of Science, and ProQuest up to March 10, 2025, identifying 775 articles. After removing duplicates, 358 articles were screened, 28 underwent full-text review, and 21 studies were included, with 18 providing sufficient metrics for meta-analysis. The MIDAS module in Stata was used to pool sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC) for the best-performing models, assessed via a bivariate mixed-effects model. Risk of bias was evaluated using PROBAST + AI.

Results

The 21 studies included 261,477 patients (mean age 58.74 years, 66.67% male), predominantly from China (11 studies). Imaging modalities included CT angiography (10 studies) and noncontrast/contrast-enhanced CT (five studies). The best models (18 studies, 7957 positive, 41,852 negative cases) achieved a pooled sensitivity of 0.92 (95% CI: 0.87–0.96), a specificity of 0.95 (95% CI: 0.91–0.98), and an AUROC of 0.98 (95% CI: 0.96–0.99). Deep learning models (DenseNet121, Attention U-Net) predominated. Moderate to substantial heterogeneity in both pooled sensitivity (I² = 64.85%) and specificity (I² = 62.87%), likely due to diverse ML algorithms and imaging protocols, was identified.

Conclusions

ML models show strong diagnostic performance for detecting aortic dissection, highlighting their promise for clinical use. However, the standardization of methodologies is essential to reduce variability and promote broader clinical implementation.

Abstract Image

查看原文本刊更多论文

机器学习在主动脉夹层诊断和预测中的应用：系统综述和荟萃分析

背景：主动脉夹层是一种危及生命的疾病，需要快速准确的诊断。机器学习（ML）模型在增强成像模式的诊断性能方面显示出了希望，但它们的综合功效尚不清楚。本系统综述和荟萃分析评估了ML模型检测主动脉夹层的诊断准确性。方法按照PRISMA指南，检索PubMed、Scopus、Embase、Web of Science和ProQuest，检索时间截止到2025年3月10日，共检索到775篇文章。在剔除重复项后，筛选了358篇文章，其中28篇进行了全文审查，纳入了21篇研究，其中18篇为meta分析提供了足够的指标。Stata中的MIDAS模块用于汇集最佳模型的敏感性、特异性和受试者工作特征曲线（AUROC）下的面积，并通过二元混合效应模型进行评估。使用PROBAST + AI评估偏倚风险。结果21项研究纳入261477例患者，平均年龄58.74岁，男性66.67%，主要来自中国（11项研究）。成像方式包括CT血管造影（10项研究）和非对比/增强CT（5项研究）。最佳模型（18项研究，7957例阳性，41852例阴性）的总敏感性为0.92 (95% CI: 0.87-0.96)，特异性为0.95 (95% CI: 0.91-0.98)， AUROC为0.98 （95% CI: 0.96-0.99）。深度学习模型（DenseNet121, Attention U-Net）占主导地位。在合并敏感性（I2 = 64.85%）和特异性（I2 = 62.87%）中发现了中度至重度异质性，可能是由于不同的ML算法和成像方案。结论ML模型对主动脉夹层的诊断效果较好，具有较好的临床应用前景。然而，方法的标准化对于减少可变性和促进更广泛的临床实施至关重要。

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

Journal of Cardiac Surgery 医学-外科

CiteScore

2.90

自引率

12.50%

发文量

976

审稿时长

3-8 weeks

期刊介绍： Journal of Cardiac Surgery (JCS) is a peer-reviewed journal devoted to contemporary surgical treatment of cardiac disease. Renown for its detailed "how to" methods, JCS''s well-illustrated, concise technical articles, critical reviews and commentaries are highly valued by dedicated readers worldwide. With Editor-in-Chief Harold Lazar, MD and an internationally prominent editorial board, JCS continues its 20-year history as an important professional resource. Editorial coverage includes biologic support, mechanical cardiac assist and/or replacement and surgical techniques, and features current material on topics such as OPCAB surgery, stented and stentless valves, endovascular stent placement, atrial fibrillation, transplantation, percutaneous valve repair/replacement, left ventricular restoration surgery, immunobiology, and bridges to transplant and recovery. In addition, special sections (Images in Cardiac Surgery, Cardiac Regeneration) and historical reviews stimulate reader interest. The journal also routinely publishes proceedings of important international symposia in a timely manner.