Machine Learning Detects Left Ventricular Hypertrophy
A systematic review and meta-analysis registered CRD42024617183 pooled 25 studies through November 12, 2025 that evaluated machine learning models for detecting left ventricular hypertrophy (LVH). Pooled results: ECG-based models sensitivity 0.76 (95% CI 0.66–0.84) and specificity 0.84 (95% CI 0.78–0.89); clinical-feature models sensitivity 0.78 and specificity 0.71; echocardiography-based models sensitivity 0.71–0.94 and specificity 0.67–0.96, though extreme heterogeneity limits confidence and necessitates further imaging-focused research.
Key Points
- 1Report pooled diagnostic accuracy across 25 studies: ECG sensitivity 0.76, specificity 0.84.
- 2Highlight heterogeneity and limited evidence undermining pooled estimates' reliability for clinical deployment.
- 3Recommend focus on high-quality imaging-based model development and rigorous external validation.
Scoring Rationale
Systematic meta-analysis gives useful pooled accuracy estimates but extreme heterogeneity and limited evidence reduce confidence.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
