**5. Conclusions and Future Work**

The main motivation of this work is enabling the high-quality real-time monitoring of CRPs at the homes of patients, designed and supervised remotely by the cardiac rehabilitation team. For this, we have proposed: (1) integrating a high-quality protocol based on clinical guidelines for monitoring the HR of the patient in a personalized way; (2) providing a real-time monitoring during sessions using a wearable wrist-worn device with heart rate sensor; and (3) using a wearable-mobile-cloud platform for collecting and synchronizing data between patients and the cardiac rehabilitation team.

The methodology of this work has been focused on modeling the theoretical approaches for developing an wearable application for real-time monitoring using wrist-worn devices. In order to address this challenge, first, a fuzzy model is proposed to describe under a linguistic approach the heart rate stream by means of three representative terms: *low, adequate, and high*. Fuzzy modifiers and fuzzy temporal windows are included in the methodology. The fuzzy approach provides a flexible evaluation of the HR stream: (1) enabling a intuitive real-time monitoring in wrist-worn wearable devices during sessions; and (2) providing visual and gradual advice, for which intensity is related to the degree of the terms.

On the other hand, an evaluation of fuzzy modifiers and fuzzy temporal windows is included to generate more accurate and flexible terms. In Section 4, the impact of fuzzy modifiers and temporal windows over the linguistic terms is analyzed by means of surveys based on cases. They have been evaluated by the cardiac rehabilitation team from the Hospital Complex of Jaen (Spain), indicating the most appropriated semantics for each linguistic term.

In future works, the approach will be extended to generate linguistic recommendations and summaries of the further sessions for the patients and the cardiac rehabilitation team. In this work, we have introduced the aggregation as a straight indicator in a 4-star scale, but a further analysis of the heart rate stream will provide intelligent and automatic feedback for cardiologists to detect weak points in the sessions of patients.

**Acknowledgments:** This contribution has been supported by the project PI-0203- 2016 from the Council of Health for the Andalucian Health Service, Spain together the research project TIN2015-66524-P from the Spanish government.

**Author Contributions:** J.M.Q. designed the methodology and wrote the paper; M.R.F.O. designed the rehabilitation protocol and led the experiments; M.D.P.A. contributed materials and analysis tools; M.E.E. analyzed and improved the methodology and the paper.

**Conflicts of Interest:** The authors declare no conflict of interest.
