Immersive Virtual Reality Using a Head Mounted Display and Modelling Using Machine Learning Algorithms to Assess Risk of Falling in the Elderly and Patients With Parkinson's Disease.

The process of ageing affects at the same time the sensory, cognitive and driving functions. Furthermore, ageing is often accompanied by pathologies increasing the effects of the senescence. An ageing subject will have then more difficulties in maintaining balance control and will have a falling risk with sometimes critical consequences for the quality of life.

The risk of fall is estimated by tests at the same time of current life and with scores of sensitivity and specificity which must be improved. In a review including 25 studies (2 314 subjects), show a sensitivity of 32 % and a specificity of 73 % on the test "Timed Up and Go" (TUG) with a threshold at 13.5 seconds.

In addition, the fall occurs in a multifactorial context when a subject interacts with his environment. It therefore seems essential to test balance control or falling risk of individuals as close as possible to the situations of daily life. This research, based on the TUG, will aim to assess the neuro-psycho-motor behavior of subjects in situations close to daily life using a Virtual Reality (VR) and Human Metrology platform.

The results could ultimately lead to increased sensitivity and specificity in assessing the risk of falling with a TUG performed in VR, compared to the classic TUG, which is commonly used by healthcare professionals and thus allow for earlier or more appropriate management of the subject in preventing the risk of falling. This could allow healthcare professionals to better understand the risk of falling and thus guide medical recommendations and prescribing, particularly in terms of appropriate physical activity programs.