| The incidence of Alzheimer is more than6.6%in the over65s. Alzheimer maycause physical decline, memory decrease, daily living skills decline, bedridden oreven death. So Alzheimer will seriously impact the health and the quality of life ofelder. The purpose of this study is to use wireless sensor network and multimediatechnologies to develop a cognitive function analysis system. This system can notonly assist long-term care institutions to create new social activities for elder, but alsotrain their cognitive abilities to reduce the risk of Alzheimer.The system consists of three parts:(1) Data acquisition: With Fun Cube as the maininteractive tools, we will collect participants’ operation steps, error count and finishtime and transfer these information to the clouds server, using multimedia technologypresent the content and the rules of the game. According to the orientation, workingmemory, attention and processing speed cognitive abilities, we designed andimplemented a total of11kinds of games.(2) Data analysis: The clouds server willfirstly match the collected data string, screening whether there is any virus. Then theclouds server will calculate and store the quantitative results of mental indicators. Weuse Aho-Corasick algorithm as the clouds string comparison algorithm and usemultithreading technology to speed up the efficiency of comparison.(3) Datafeedback: Establish a web physician can access to that can present participants’operation steps and automatic screen high risk group who have cognitive dysfunction,so that the doctor can start care plan early. We use Visual Studio software to designweb which can extract and analyze the operation time, operation steps and correctnessin storage.After the success of the system construction, we processed three experiments: thefirst experiment is the analysis of this system’s reliability and validity: to evaluate theconnection between this system evaluation results and the traditional printedevaluation results; the second experiment is the survey of this system’s satisfaction:use Technology Acceptance Model2questionnaire to test the elder’ science andtechnology accept degree of this system. The above two experiments proceed inChang Gung Health and Culture Village. The experiment is divided into four steps:(1)Pretest: participants sign the respondent’s consent letter, and test MMSE and FABquestionnaire which are as the judgment condition of grouping.(2) Grouping: testersrandomly divided into two groups (group A and group B) and we use MMSE score totest whether there is no difference in two groups (i.e., p>0.05). If there is a difference,we artificially adjust to indifference.(3) The first test: group A does MoCA testing;Group B does Fun Cube mental games and TAM2questionnaire.(4) The second test:two weeks later, Group A and Group B exchange. We have recruited43elder to testthis system, there are37people adhere to the finished the whole test which27peopleare women and10people are man. The testers’ average age is77.9. The testers’average length of education is14.6and the average score in MMSE is27.0point. Inthe system’s reliability and validity, we found that the Pearson correlation coefficientbetween this system evaluation results and traditional printed evaluation results was0.78, belongs to the strong correlation, which show that our system have a certaincredibility. In the survey of this system’s satisfaction, we found that the score ofintention to use is above5.6point, the score of perceived usefulness is above5.33point, the score of perceived ease of use is above5.7point, the score of entertainmentis above5.88point, the score of subjective impression is above5.56point, the score of job relevance is above5.12point, the score of output quality is above5.88point,the score of result demonstrability is above3.86point, the score of voluntariness isabove5.98point and the score of preference is4.77point, which explain that oursystem is funny, simple, easy to operate, and has a strong attraction for elder. Thethird experiment is the performance analysis of Aho-Corasick algorithm in serverusing multithreading technology. This experiment is run on the HP9000Superdomeserver. The results showed that: two CPU can promote about50.2%performance andfour CPU can promote about74.3%, relative to one CPU.This research combined wireless sensor network and multimedia technologies forcognitive function assessment. On the one hand, we developed games related totraditional printed scales, which improves the shortcoming of traditional printedscales and can assist doctors to discover people who has Alzheimer as soon aspossible. On the other hand, we exploited and implement some cognitive functiontraining games, through which can exercise and improve the elder’ cognitive functionto delay the occurrence and development of the Alzheimer’s disease. |
PDF Full Text Request