Cardiovascular diseases represent the leading cause of death worldwide. Thus, cardiovascular rehabilitation programs are crucial to mitigate the deaths caused by this condition each year, mainly in patients with coronary artery disease. COVID-19 was not only a challenge in this area but also an opportunity to open remote or hybrid versions of these programs, potentially reducing the number of patients who leave rehabilitation programs due to geographical/time barriers. This paper presents a method for building a cardiovascular rehabilitation prediction model using retrospective and prospective data with different features using stacked machine learning, transfer feature learning, and the joint distribution adaptation tool to address this problem. We illustrate the method over a Chilean rehabilitation center, where the prediction performance results obtained for 10-fold cross-validation achieved error levels with an NMSE of $0.03\pm 0.013$ and an $R^2$ of $63\pm 19\%$, where the best-achieved performance was an error level with a normalized mean squared error of 0.008 and an $R^2$ up to $92\%$. The results are encouraging for remote cardiovascular rehabilitation programs because these models could support the prioritization of remote patients needing more help to succeed in the current rehabilitation phase.

Background: Preeclampsia remains a leading cause of maternal morbidity worldwide. There is a critical need for predictive systems that not only perform accurately but also provide interpretable insights for clinical decision-making. This work introduces SK-MOEFS, an explainable framework based on fuzzy logic and multi-objective evolutionary optimization, designed to classify preeclampsia risk while generating clinically interpretable rules. Methods: The model integrates fuzzy decision trees with a genetic algorithm to identify a compact and relevant set of rules, optimized for both accuracy and interpretability. The system was trained and evaluated on third-trimester pregnancy data from a publicly available, multi-ethnic cohort comprising 574 individuals. All processes, including preprocessing, training, and evaluation, were conducted using open-source tools, ensuring reproducibility. Results: SK-MOEFS achieved 91% classification accuracy, an AUC of 0.89, and a recall of 0.88—outperforming other standard interpretable models while maintaining high transparency. The model emphasizes minimizing false negatives, which is critical in clinical risk stratification for preeclampsia. Conclusions: Beyond predictive performance, SK-MOEFS offers a rule translation and defuzzification layer that outputs probabilistic interpretations in natural language, enhancing its suitability for clinical use. This framework provides an effective bridge between algorithmic inference and human clinical judgment, supporting transparent and reliable decision-making in maternal care.