VR training for heart patients using an AI solution

At the beginning of this project, there was no motivational training interventions that were simple and effective and did not require supervision. 

The project's solution was to reuse and transform the hospital's existing training equipment, like stationary bicycles, for sensory-stimulating virtual nature experiences. It would not require WiFi and only one click was needed to start the training experience.

Aim

This project sought to develop an AI algorithm to be built into SYNCSENSE's VR solution that could adapt to the physical capabilities of the individual heart patient, thus making the VR training intelligent and efficient. The AI solution was to automatically identify training equipment and syncronise the speed of the bicycle to the speed in the VR experience.

The project had three phases:

1. Development of the AI algorithm
2. Migration of the AI algorithm for the VR solution
3. Testing and adapting the new AI based VR solution in a clinical setting

Results

The developed solution has an AI-based sensor that fits any exercise equipment that has a movable component. The AI runs a one second, three-stage loop in order to identify the exercise equipment and adapt the speed of it.

The developed VR solution offers both active and passive VR training for heart patients. Active training takes place in combination with training equipment, like stationary bicycles, and sensors and can work as motivation for training, continued training, and can also facilitate self-training. Passive training takes place without training equipment and sensors and can help calm patients and work as a recreation activity.

Partners

Partners on the project included SYNCSENSE, who undertook development and delivery of the solution, as well as training of the algorithm. Maersk Mc-Kinney Moller Institute at the University of Southern Denmark provided counseling in regards to the AI algorithm, while CAI-X was in charge of project management and the implementation strategy.

Clinical testing took place at the Department of Cardiology at Odense University Hospital.

External funding

The project was funded by Innovation Fund Denmark.