JPBS

Background: The method to understand the “now” behavior of dementia patients is based on direct behavioral observations. However, in Japan there are a limited number of trained professionals capable of performing direct observations. This study aimed to classify patients with dementia into high and low mental function groups on the basis of the cognitive, motivational, and emotional functions of the Mental Function Impairment Scale, and to visualize the behavioral patterns in each group using sensing technology as an alternative to direct observations. Methods: We classified participants’ cognitive, motivational, and emotional functions into high and low mental function groups and analyzed the characteristics of each groupʼs behavioral patterns using a network analysis based on graph theory. Location information collected by wearing a sensor wristband (Fujitsuʼs ubiquitous software) was used to analyze behavioral patterns, and the Mental Function Impairment Scale was used to assess cognitive, motivational, and emotional functions. Results: The behavior of those with low cognitive functioning showed a uniform behavioral pattern, and those with low motivational and emotional functioning were considered to be in a state where the behavior itself could not be generated. Conclusions: The sensor technology accurately reflected the behavioral patterns of the high- and low-value groups across each of the three MENFIS sub-items, cognitive function, motivational function, and emotional function. Individuals with low cognitive function showed a uniform, stationary behavioral pattern, indicative of executive function deficits.

graph theory; dementia; pattern of behavior