Healthcare researchers have developed an online, interpretable machine learning system that can predict cardiometabolic multimorbidity (CMM) risk in patients with heart disease. type 2 diabetes (T2DM), which could change early intervention strategies.
Machine learning identifies high-risk patients for type 2 diabetes
CMM is defined as the co-occurrence of cardiovascular disease, diabetic complications, and metabolic disorders and significantly increases mortality and healthcare burden in the T2DM population. Early identification of high-risk individuals is important for targeted prevention and management. Xiaohan Liu and his team designed an AI-driven predictive model to address this challenge.
The researchers used data from 793 patients from a tertiary hospital, dividing participants into a training (80%) set and an internal validation (20%) set, with an additional 360 patients from an independent center for external validation. Recursive feature removal using a random forest algorithm identified nine key predictors. Six machine learning algorithms were trained, and the stacking model performed the best. We achieved an area under the curve of 0.868 in internal validation and maintained robust performance with an area under the curve of 0.822 in external validation.
Interpretability of the model was ensured by SHapley Additive explanations and Local Interpretable Model-Agnostic Explanations, allowing clinicians to understand the contribution of individual risk factors.
Online tools support clinical decision making
The system is now available online and has the potential to help healthcare providers quickly assess CMM risk in T2DM patients and implement timely interventions to slow disease progression. This online tool bridges the gap between complex AI models and real-world clinical use, supporting personalized care decision-making.
Machine learning has the potential to enhance risk stratification, but limitations of this study include reliance on data from specific hospital populations, which may limit generalizability. Further multicenter trials are needed to validate the model’s performance across diverse ethnic and demographic groups.
As the prevalence of diabetes increases globally, integrating AI-based risk prediction tools has the potential to improve outcomes for millions of people at risk of cardiometabolic complications. This study shows how artificial intelligence in healthcare can be leveraged for precision medicine to support clinicians in making data-driven interventions.
reference
Liu X et al. An online interpretable machine learning model for predicting cardiometabolic multimorbidity risk in patients with type 2 diabetes. Science representative 2026; Doi:10.1038/s41598-026-36923-2.
