This study explores the potential of using machine learning algorithms to evaluate heart failure through continuous monitoring of non-invasive signals, specifically photoplethysmogram (PPG) and electrocardiogram (ECG). The research leverages data from the MIMIC-III database, which includes 682 heart failure patients and 954 controls. Key features such as QRS interval, RR interval, augmentation index, heart rate, systolic and diastolic pressure, and peak-to-peak amplitude were selected for their clinical relevance and ability to capture cardiovascular dynamics. These features were used to train machine learning models, resulting in high accuracy (98%), sensitivity (97.60%), specificity (96.90%), and precision (97.20%). The integration of PPG and ECG signals demonstrated superior performance compared to single-signal strategies, highlighting the potential for early and precise heart failure diagnosis. The study also emphasizes the importance of continuous monitoring with wearable technology, suggesting significant progress in non-invasive cardiovascular health assessment. The proposed approach holds promise for implementation in hardware systems to enable continuous monitoring, aiding in early detection and prevention of critical health conditions. The integration of PPG and ECG signals enhances the reliability of the monitoring system by providing complementary information, reducing the risk of false negatives or positives. The study's results demonstrate the effectiveness of combining these signals for heart failure assessment, with the Random Forest model achieving the highest accuracy (98.00%) and AUC (97.70%). The findings suggest that the integrated approach offers significant improvements in diagnostic accuracy and has the potential to transform cardiac healthcare. The study also highlights the clinical relevance of the approach, emphasizing its potential for early, precise, and non-invasive diagnosis of heart failure.This study explores the potential of using machine learning algorithms to evaluate heart failure through continuous monitoring of non-invasive signals, specifically photoplethysmogram (PPG) and electrocardiogram (ECG). The research leverages data from the MIMIC-III database, which includes 682 heart failure patients and 954 controls. Key features such as QRS interval, RR interval, augmentation index, heart rate, systolic and diastolic pressure, and peak-to-peak amplitude were selected for their clinical relevance and ability to capture cardiovascular dynamics. These features were used to train machine learning models, resulting in high accuracy (98%), sensitivity (97.60%), specificity (96.90%), and precision (97.20%). The integration of PPG and ECG signals demonstrated superior performance compared to single-signal strategies, highlighting the potential for early and precise heart failure diagnosis. The study also emphasizes the importance of continuous monitoring with wearable technology, suggesting significant progress in non-invasive cardiovascular health assessment. The proposed approach holds promise for implementation in hardware systems to enable continuous monitoring, aiding in early detection and prevention of critical health conditions. The integration of PPG and ECG signals enhances the reliability of the monitoring system by providing complementary information, reducing the risk of false negatives or positives. The study's results demonstrate the effectiveness of combining these signals for heart failure assessment, with the Random Forest model achieving the highest accuracy (98.00%) and AUC (97.70%). The findings suggest that the integrated approach offers significant improvements in diagnostic accuracy and has the potential to transform cardiac healthcare. The study also highlights the clinical relevance of the approach, emphasizing its potential for early, precise, and non-invasive diagnosis of heart failure.