AI prediction models with omics data utilization for atherosclerosis: A systematic scoping review and meta-analysis

Area of Science:

• Cardiovascular omics research
• Artificial intelligence in precision medicine
• High-dimensional data interpretation

Background:

• Cardiovascular disease (CVD) is a leading global cause of mortality.
• Advanced methods are needed to understand CVD pathophysiology.
• Artificial intelligence (AI) applied to omics data offers potential for precision medicine.

Purpose of the Study:

• To systematically review AI technologies in cardiovascular omics research.
• To compare AI prediction models (APMs) against conventional risk prediction models (CRMs) for atherosclerosis.

Main Methods:

• Two-phase systematic review: scoping review (218 studies) and meta-analysis (38 studies).
• Meta-analysis quantified incremental performance of APMs over CRMs using difference in area under the curve (ΔAUC).
• Focus on atherosclerosis-specific studies.

Main Results:

• Significant growth observed in AI modeling and omics methodologies.
• APMs significantly outperformed CRMs (pooled ΔAUC = 0.0586, p < 0.0001).
• Moderate heterogeneity (I² = 40.02%) observed in meta-analysis.

Conclusions:

• AI prediction models demonstrate a robust performance advantage over conventional models.
• Performance benefit of APMs was consistent across various experimental designs and validation strategies.
• AI holds significant promise for advancing cardiovascular precision medicine through omics data.