Scholarlink

Background <p>Accurate cardiovascular disease (CVD) risk prediction is critical for preventive decision-making. While traditional tools like the Framingham Risk Score (FRS) and Coronary Artery Calcium Score (CACS) are widely used, newer artificial intelligence (AI)-based models such as AICVD and Reti-CVD offer promising alternatives. However, comparative data across traditional, radiological, and AI-based tools remain limited.</p> Objective <p>This study aimed to compare the predictive efficacy of traditional, radiological, and AI-driven CVD risk tools through a network meta-analysis.</p> Methods <p>A systematic search of PubMed, Embase, Cochrane Library, and Google Scholar identified four observational studies comparing predictive tools in adults for ischemic CVD outcomes. Tools assessed included FRS, QRISK3, CACS, CIMT, baPWV, AICVD, and Reti-CVD. Primary outcomes were risk stratification accuracy and the incidence of ischemic CVD events during follow-up (up to 11&#xa0;years). A network meta-analysis using random effects models was performed in R.</p> Results <p>A total of four observational studies with 53,641 participants were included. AICVD demonstrated the highest predictive efficacy, with an 86% higher relative risk of identifying ischemic CVD compared to QRISK3 (RR: 1.86; 95% CI: 1.09–3.18). The combined use of CACS and FRS improved predictive accuracy (RR: 1.50), while CACS alone showed a modest benefit (RR: 1.29). Reti-CVD demonstrated comparable performance to CACS (RR: 0.87; 95% CI: 0.46–1.65), offering a non-invasive alternative without radiation exposure. Ranking analysis indicated AICVD &gt; CACS + FRS &gt; CACS &gt; Reti-CVD &gt; baPWV &gt; CIMT &gt; FRS &gt; QRISK3.</p> Conclusion <p>AI-based tools such as AICVD and Reti-CVD may outperform traditional methods in predicting ischemic CVD. Despite limited data, these findings highlight the potential of AI in risk prediction, warranting further validation across diverse populations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative Efficacy of Cardiovascular Risk Prediction Tools: A Network Meta-analysis of Traditional, Radiological, and AI-Based Approaches

  • 期刊论文

  • 发表日期：

    2025年9月30日

查看全文
  • Tirth Bhavsar,
  • Shashi Mundhra,
  • Rahul Kamboj,
  • Aashwin Kaushal,
  • Mamta Kamboj,
  • Kumari Uthayakumar,
  • Ajay Singh,
  • Sanjiya Arora,
  • Sachin Mahendrakumar Chaudhary,
  • Devendra Tripathi,
  • Pranav Kumar Sharma,
  • Manju Rai

摘要

Background

Accurate cardiovascular disease (CVD) risk prediction is critical for preventive decision-making. While traditional tools like the Framingham Risk Score (FRS) and Coronary Artery Calcium Score (CACS) are widely used, newer artificial intelligence (AI)-based models such as AICVD and Reti-CVD offer promising alternatives. However, comparative data across traditional, radiological, and AI-based tools remain limited.

Objective

This study aimed to compare the predictive efficacy of traditional, radiological, and AI-driven CVD risk tools through a network meta-analysis.

Methods

A systematic search of PubMed, Embase, Cochrane Library, and Google Scholar identified four observational studies comparing predictive tools in adults for ischemic CVD outcomes. Tools assessed included FRS, QRISK3, CACS, CIMT, baPWV, AICVD, and Reti-CVD. Primary outcomes were risk stratification accuracy and the incidence of ischemic CVD events during follow-up (up to 11 years). A network meta-analysis using random effects models was performed in R.

Results

A total of four observational studies with 53,641 participants were included. AICVD demonstrated the highest predictive efficacy, with an 86% higher relative risk of identifying ischemic CVD compared to QRISK3 (RR: 1.86; 95% CI: 1.09–3.18). The combined use of CACS and FRS improved predictive accuracy (RR: 1.50), while CACS alone showed a modest benefit (RR: 1.29). Reti-CVD demonstrated comparable performance to CACS (RR: 0.87; 95% CI: 0.46–1.65), offering a non-invasive alternative without radiation exposure. Ranking analysis indicated AICVD > CACS + FRS > CACS > Reti-CVD > baPWV > CIMT > FRS > QRISK3.

Conclusion

AI-based tools such as AICVD and Reti-CVD may outperform traditional methods in predicting ischemic CVD. Despite limited data, these findings highlight the potential of AI in risk prediction, warranting further validation across diverse populations.