Creating adapted environments: enhancing accessibility in virtual reality for upper limb rehabilitation through automated element adjustment

In the last decade, Virtual Reality (VR) has emerged as a promising tool for upper limb rehabilitation, effectively complementing conventional therapies. However, one of the main challenges lies in designing virtual environments that adapt to the specific needs of each patient, considering their unique motor limitations. An inadequately adapted environment can result in overexertion and the inability to perform exercises, negatively affecting both the patient's motivation and their recovery. This article hypothesizes that automatic calibration and dynamic object adjustment algorithms in virtual reality environments improve accessibility and efficiency in upper limb rehabilitation exercises for patients with SCI. For this purpose, we present an innovative calibration method that individually identifies and maps motor limitations on the left and right sides of the body. As a result, an irregular volume, formed by the interconnection of three elliptical shapes, is generated that envelops the patient and represents their safe range of movements. Furthermore, a second method is introduced that automatically readjusts the location of objects within the virtual environment to the safe space generated, optimizing the patient's accessibility and interaction with therapy elements. To test the results, an immersive VR environment was designed in which the aforementioned methods were applied for the automatic placement of virtual elements in the peripersonal space (PPS) of the participants. Testing has been carried out at the Hospital Nacional de Parapl & eacute;jicos in Toledo (HNPT) with patients suffering from spinal cord injuries (SCI) and healthy participants who are SCI specialists. The quantitative results obtained demonstrate that this dynamic adjustment of the environment allows for adaptation that leads to a 100% success rate in task completion after the automatic adjustment, compared to a 62.5% success rate when using a configuration with virtual elements adapted to the motor capabilities of a healthy person (for both healthy participants and patients). This adjustment not only facilitates a greater number of exercise repetitions, but also reduces the time needed to access each object, with an average reduction in time of 47.94% across the entire sample. This reduction is even more significant when considering only the group of SCI patients, with a reduction of 53.78%. Additionally, the qualitative evaluation complements the study with a perception of ease of use for the calibration (mean = 1.29 +/- 0.46) and low complexity in accessing the interactive objects after the automatic adjustment (mean = 1.12 +/- 0.45). These results demonstrate the effectiveness of the proposed algorithms and the improved user experience.