Establishment and validation of a machine learning-based predictive model for in-hospital mortality risk in acute myocardial infarction patients complicated with diabetes mellitus - Summary - MDSpire
Advertisement
Establishment and validation of a machine learning-based predictive model for in-hospital mortality risk in acute myocardial infarction patients complicated with diabetes mellitus
To develop and validate a machine learning-based predictive model for in-hospital mortality risk in patients with acute myocardial infarction complicated by diabetes mellitus, addressing a critical gap in clinical risk assessment.
Key Findings:
Seven predictors were identified: heart rate, neutrophil count, monocyte count, NLR, serum albumin, total bilirubin, and urea nitrogen, which are critical for assessing mortality risk.
The XGBoost-based model achieved the highest AUC and was selected as the preferred model, indicating its superior predictive capability.
The SHAP method enhanced the interpretability and clinical credibility of the mortality prediction model, facilitating its integration into clinical practice.
Interpretation:
The developed machine learning models can aid in identifying high-risk patients and implementing early intervention measures, potentially reducing in-hospital mortality rates.
Limitations:
The model is limited to clinical settings similar to the study cohorts and is not intended for broad clinical application, which may affect its generalizability.
Conclusion:
Interpretable machine learning models have been developed to predict in-hospital mortality risk in AMI patients with DM, providing insights into the influence of various features on prediction outcomes and highlighting the need for further validation in diverse clinical settings.
Researchers found that patients with higher waist circumference and lower grip strength had the greatest risk for developing type 2 diabetes during long-term follow-up.
