A Virtual Reality Memory Test for Assessing Visuospatial Memory

Abstract

AbstractBackgroundVisuospatial memory impairment is one of the common symptoms of Alzheimer’s disease. However, impairment in visuospatial and visuospatial memory function has not been focused on as much as that of episodic memory function. We developed a visuospatial memory test, the Hidden Objects Test (HOT), that was designed for people to go around the living room of a house created by virtual reality (VR) and find hidden objects.MethodThis study group comprised 17 Alzheimer’s disease dementia (ADD), 14 amnestic mild cognitive impairment (aMCI), and 15 normal controls (NC). All participants completed a sandardized neuropsychological battery and magnetic resonance imaging (MRI). Out of 46 participants, 28 received amyloid positron emission tomography (PET). Participants performed a novel visuospatial memory test in a virtual environment using VR headset HTC Vive. Walking trajectory in the virtual environment was tracked with 13Hz sampling rate. Participants’ movement path was basically assessed by the total distance, duration, and mean speed.ResultTotal HOT score and five subtests scores were differed among ADD, aMCI, and NC (p <0.001). We performed walking trajectory pattern mining as well as basic trajectory feature analysis such as total distance, duration, and speed. Total duration was significantly greater in ADD than in NC (p = 0.008). The number of outliers, over 95% of estimated trajectory, was significantly higher in ADD than in NC (p = 0.002). The number of stay points, an index that participants stayed in the same position, were significantly higher in ADD and aMCI patients compared to NC (ADD vs. NC: p = 0.003, aMCI vs. NC: p = 0.019). Walking trajectory analysis suggested that ADD and aMCI patients wandered rather than going straight to the hidden objects.ConclusionOur study showed that HOT total scores differed among ADD, aMCI, and NC groups. In addition, our study provided application of trajectory data mining methods to explore the patterns of ADD and aMCI patients' walking trajectory, which could be applied to detect behavior of the patients in environments equipped with the internet of things such as future smart home or facilities.