Enhancing Heart Failure Management: Integrating AI with Telemonitoring Systems for Advanced Alarm-Based Solutions

Heart failure (HF) poses significant global health challenges, necessitating advanced management strategies. Telemonitoring, coupled with artificial intelligence (AI), offers a promising approach to enhance HF care by remotely monitoring patients and predicting exacerbations. This paper introduces vitalera, a telemonitoring platform integrating AI to optimize patient management through personalized health data analysis and smart alarms. Key to our study is the improvement of vitalera's alarm-based system using machine learning models. An ongoing observational clinical study across hospitals in Spain and Romania provides data for training and validating the system. Initial results using traditional Machine Learning (ML) models evaluated by means of cross-validation show promising performance, achieving high recall rates crucial for timely detection of critical alarms and outperforming the current rule-based system. Challenges include optimizing model performance across varied hospital contexts and managing dataset imbalance. This study underscores AI's potential in revolutionizing HF telemonitoring, aiming for enhanced patient outcomes and quality of care.