Assessing accelerometric, gait and physiological parameters of induced spatial orientation in people with MCI or mild dementia and older healthy cohorts

Abstract

AbstractBackgroundPersons with dementia often experience spatial disorientation. They may benefit from the use of assistive technological devices to support outdoor navigation. One requirement for such systems is the automated detection of spatial disorientation to provide real time support. Previous studies showed that accelerometry and physiology signals could be informative in detecting phases of disorientation in real time. In an earlier field study, we developed a pattern recognition model of disorientated behavior based on accelerometric data from a real‐word environment. Here, we use the hybrid virtual and real‐world environment of the Gait Real‐Time Analysis Interactive Lab (GRAIL) (Figure one) to investigate accelerometric, gait and physiological parameters of induced spatial orientation among people with dementia and older healthy people.MethodWe implemented a 3D virtual environment (VE) of the Rostock city center into the GRAIL system, which consists of an instrumented treadmill and a 180° projection screen. Cognitively healthy older participants (age: 60‐85, MMSE ≥ 28) and PwD (age: 60‐85, MMSE ≥ 15 und ≤ 27) are included in this ongoing study. Participants walk a predefined route in the VE twice. In the first walk, they learn the route guided by the experimenter and then are asked to find the way to the goal destination unguided. Spatial disorientation is induced by manipulating landmarks in the 3D environment (for only the healthy older subjects in the experimental group), and resulting phases of spatial disorientation are video recorded for later offline annotation using a customized annotation scheme. Participants are equipped with wearable sensors which record heart rate, accelerometry and skin conductance. Afterwards, participants fill questionnaires examining usability and immersion.ResultRecruitment is currently ongoing. Four old healthy participants (mean age = 67.5 , SD = 3.69) (3 female) and one (male) PwD (age = 79) have been measured so far. All participants completed the experiment. Evaluation of questionnaire responses indicates an overall suitability of our setup. Further results will be presented at the conference.ConclusionOur study serves as a basis for the development of situation‐aware sensor‐based ATD that support PwD in wayfinding. This study contributes towards improving independent mobility among PwD.