Using an AI-powered VR-coach to help clients do exercises without a medical professional
Duration: 01/2022 – 12/2023 Funding: MIT
There is growing clinical evidence that patients in better physical and mental shape respond better to medical treatments compared to their peers. Often, the patient’s condition prior to a medical intervention has a great effect on the results of the intervention, however such prevalidation (in contrast with revalidation) is not always done prior to a medical intervention. One of the reasons is that it often depends on the patient’s willingness to do certain exercises, and the limited clinical supervision.
In PREHELIOS, the goal is to create an AI-driven VR coach, who is able to guide patients through exercises for prevalidation (prior to treatment) and revalidation (post-treatment). Examples of such exercises include physical therapy exercises, as well as mental exercises which may prevent the occurrence of postoperative delirium, particularly in older adults. Patients will be able to use the VR-based solution in their own home, with minimal clinical supervision.
Almende will provide the knowledge and expertise on the AI side, and will use this to develop the virtual AI coach which will then be integrated into a VR environment. The coach will respond to triggers obtained from the VR goggles and peripherals to determine whether an exercise is performed correctly, and will then suggest corrections to the patient based on several behavioural theories.
Almende has been working on digital and/or AI-driven coaching for several years, notably in projects like KRISTINA. We see applications in using (partially) automated coaching in our spin-offs active in healthcare or communications, to reduce the communication burden currently experienced by e.g. healthcare workers.