**5. Conclusion**

We presented in this paper the technical aspects and the measured performance of MedBike. Developing MedBike around Unity 3D has been an excellent design choice because of its flexibility and ease of programming new functions. In addition, Unity 3D is one of the video game industry standards and the development of

additional games is relatively easy. During the pilot project in the patient's home, we were able to demonstrate that the system can work on a standard home internet connection. Before the medical pilot study, we performed numerous (40 persons) usability studies in our laboratory to optimize the interface design. Over 90% of the users of MedBike were very impressed by its performance and would certainly like to use it capabilities for their home exercise program if the system was not so expensive. The main reason for MedBike high-cost is because we use the KICKR from Wahoo and a relatively high-end mountain bike making the average price of the bike to be around \$5 K CAN. By using low-cost stationary bikes (\$265), a new low-cost (\$300) resistance control system based on stepper motor controlling a simple bike, and a SONY Play station 4 (\$350) instead of a graphic PC, we were able to design a low-cost system that will be below \$1500.

The current version of MedBike has shown to be effective for delivering exercise programs to post procedural and event cardiac patients, saving time, encouraging adherence to the exercise program, and ensuring patient safety by direct clinician monitoring. We also believe that by adding a machine learning functionality that will analyze the MedBike biometric data in real-time to determine if the exercise level fits in the bounds of the prescribed exercise program safely, the system can be extensively used in sports facilities or at home to encourage higher activity levels in an older and detrained population.
