The ErgoRehab proposal consists of software based on deep-tech algorithms for evaluation of how a movement is performed. These tools have been validated for healthy athletes and implemented for personalized training and provided 25% more benefits compared to best practices.
The ErgoRehab software has two key inputs.
The first inputs are generated from a milder version of the movement evaluation tests implemented for athletes. Three groups of tests are employed focusing on dynamic movement, static endurance, dynamic endurance / recovery. Movement algorithms take the test results and provide movement metrics that completely characterize how a resistance is moved and these are employed as inputs to the ErgoRehab software.
The second key input to the ErgoRehab software has to do with bone mineral density. Aside from the average BMD and the T-score a metric related to spatial variation of the BMD is employed. In the first iteration of the ErgoRehab software the TBS score from TBS iNsight is used.
With these data the ErgoRehab software generates assessments of unbalanced movement and weak points which together with the TBS score can provide an imporved estimate on fracture risk and recommendations for physical therapy modifications.
The underlying principle is that proper movement patterns lead to more uniform stress on the bone and therefore more uniform remineralization and healing. Also, risk of fracture is not just a function of average BMD but also bone microstructure - this is well known in the literature. Finally, it is hypothesized that osteopenia and microstructural non-uniformities can indirectly affect movement by various protective mechanisms (musculoskeletal and/or neuromuscular).
ErgoRehab could be employed during the recovery stage, post-fracture, to provide a more personalized and patient-centric recovery experience.
We have participated in the preparation of two recent EU Horizon2020 proposals that were related to rehabilitation (hand injuries and exoskeletons for hip injuries) where our contribution was deep monitoring of patients during the recovery period and providing more intelligent input data for the AI systems involved. We are currently investigating the implementation in rehabilitation applications mostly as a means for in-depth monitoring or recovery.
With our advanced movement evaluation and advanced algorithms we anticipate the following:
1. more effective healing of both bone and muscle and movement functional capability (i.e., recovery of an increased % of previous capability).
2. more efficient healing in terms of time and necessary resources.
3. fewer problems long-term, e.g., pain or re-injuries.
Important to note that our algorithms are based on first-principles without getting lost in joint-by-joint biomechanics. As such, we do not need any data training period to train an AI as we immediately know what the data mean.
Consequently, we can proceed to a pilot study very fast.
However, in order to connect to other sources of data, e.g., TBS and long-term outcomes, a ML approach would be beneficial and is considered a part of "version 2" or ErgoRehab.