This is a page that curates AI-related papers published worldwide. All content here is summarized using Google Gemini and operated on a non-profit basis. Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.
Machine Learning Solutions Integrated in an IoT Healthcare Platform for Heart Failure Risk Stratification
Created by
Haebom
Author
Aiman Faiz, Anna Maria De Roberto, Claudio Pascarelli, Gianvito Mitrano, Gianluca Fimiani, Marina Garofano, Genoveffa Tortora, Mariangela Lazoi, Claudio Passino, Alessia Bramanti
Outline
This paper presents a machine learning (ML)-based predictive model for identifying patients at risk of chronic heart failure (HF). The model is an ensemble learning approach with a modified stacking technique that uses two expert models that leverage clinical and echocardiographic features and a meta-model that combines their predictions. Evaluation results using a real dataset show that it effectively identifies patients at risk of HF, achieving high sensitivity (95%) and moderate accuracy (84%). The model will be used to select participants for the remote monitoring program of the PrediHealth research project, and outperforms existing baseline models.
Takeaways, Limitations
•
Takeaways:
◦
These results suggest that ML-based risk stratification models can serve as valuable decision support tools to aid in early identification and personalized management of patients at risk for heart failure.
◦
It shows that it can be effectively utilized in remote monitoring programs such as the PrediHealth project.
◦
Combining clinical and echocardiographic features, we demonstrate improved performance over existing simple feature-based models.
•
Limitations:
◦
Accuracy (84%) may be considered moderate in some ML contexts.
◦
Further validation of the size and generalizability of the study dataset is needed.
◦
Additional performance evaluations in diverse environments and populations are needed.
