Meta-analysis of 13 studies in peripheral arterial disease: machine learning models outperformed traditional regression for predicting major adverse limb events; gradient-boosted models had the highest diagnostic odds ratio (36.6) versus a lowest of 3.3 for regression.
Meta-analysisn = 13 studies
Ravindhran et al. Machine Learning Models in the Prediction of Adverse Outcomes in Peripheral Arterial Disease: Meta-Analysis