Utilización de aprendizaje automático para predecir problemas cardíacos a partir de datos de dispositivos portátiles

Cardiac problems are one of the leading causes of mortality worldwide, accounting for approximately 17 million deaths each year. Early detection of these conditions is essential to reduce mortality rates and improve the population's quality of life. The objective of this study was to develop a predictive system that, using data obtained from wearable devices, specifically smartwatches, can identify the likelihood of cardiac problems. Publicly available data were collected from the Google Fit platform and an open dataset from Kaggle. The methodology included preprocessing techniques such as imputing missing values, normalizing data, and encoding categorical variables. Four machine learning algorithms were trained and evaluated: Support Vector Machine (SVM), XGBoost, Decision Tree, and Random Forest. The performance of the models was assessed using accuracy, recall, F1-score, and area under the curve (AUC). The results showed that the Random Forest and SVM algorithms achieved the best performances, reaching an F1-score above 90% and a maximum AUC of 93%, demonstrating strong predictive capability for the early identification of cardiac problems. In conclusion, this study demonstrates that combining data from wearable devices with machine learning algorithms provides a reliable and accessible tool for the early detection and prevention of cardiac diseases, thereby contributing to the advancement of preventive medicine.