【Background】 When children suffer a sudden illness or injury, their parents cannot always determine whether the child should immediately go to a medical institution, and they may then call an ambulance. We developed a smartphone application for parents to provide emergency medical service according to the condition and urgency of pediatric patients. 【Method】 When the parent selects the appropriate symptoms displayed by the app, the urgency of the condition is determined by medical algorithms in the app. The app provides emergency medical services such as calling for an ambulance and information on hospitals that can be visited immediately depending on the urgency. This app was released in September 2015 and has been available on Google Play and the iOS App store since April 2016. 【Results】 The app was downloaded 7,780 times from September 2015 to December 2016. It has been used 11,560 times in total and has provided users with emergency medical services such as calls for ambulances 4,979 times. 【Conclusion】 This application appears useful because it was utilized intensively during the closure of a pediatric hospital. In the future, we would like to improve the accuracy of the algorithms used and analyze the data accumulated by the app.

 Students’ learning behavior in information science practice were analyzed from the viewpoints as the latency. The items you can know in potential ranks from answers of questionnaires in last year were indicated, but when the presence of structural characteristics were confirmed once again using item entropy, the items it doesn’t go along was judged as looseness of a reaction of an answer, and the former could see latency, and item entropy were the big tendency for these. The latter couldn’t see latency and item entropy were low values. Item entropy could think there were spread of dispersion of consciousness or own awareness from points as learning behavior by the big item, and this tendency didn’t change through whole practice period. The similarity in multidimensional scaling and cluster analysis were same state. It seems also to develop from these points as analysis that more learning behavior were quantitative into own others’ relationship.

 Some articles point out that there are regional gaps of medical resources in Hokkaido. The medical plan in Hokkaido has mentioned the importance of solving those maldistributions. However, there is little research on regional distribution of healthcare human resources. We evaluate the regional distributions of physicians, nurses, pharmacists, radiological technologists, physical therapists (PT), and occupational therapists (OT) using Gini coefficient, which is widely used to evaluate income inequality. The result shows that Gini coefficients are 0.137 for RT, 0.159 for pharmacists, 0.160 for physicians, 0.163 for nurses, 0.276 for PT, and 0.289 for OT. Gini coefficients for PT and OT are higher than Gini coefficients for radiological technologists, pharmacists, physicians, and nurses, which means the possibilities of there being maldistributions of PT and OT. Therefore, it is necessary to consider PT and OT supply in the regions where the shortages of the supply is likely to exist.

 Recently, home care services are growing based on the Integrated Community Care System in Japan. However, it has been suggested that service utilization is different in each community. In this paper, we examined a statistical method to analyze the usage rate of home care services at the community level using Geographic Information Systems. We prepared anonymized medical prescription data that included the gender, five-year age group, code of medical facility, and ZIP code of the patient’s address. The usage rates of home care services were calculated in each cohort and regional community. Finally, we compared usage rates with population composition, presence of bigger apartments, number of clinics, distance from clinics, presence of nursing homes for the elderly, and so on. As a result, when we mapped the usage rates of home care services and calculated the time distance, the lower usage rate areas had a tendency to appear within lower accessibility areas. In addition, it is suggested that the presence of bigger apartments and nursing homes for the elderly affects differences in usage rates of home care services. Using this method might reveal the reasons for regional differences in home care utilization.

 ［Objective］ We investigate emotional expressions in Alzheimer’s Disease (AD) patient narratives to realize AD screening. ［Method］ We collected people’s episodes through crowdsourcing and developed Japanese Inquiry and Word Count (JIWC) to analyze emotional expressions. We analyze patient’s utterance contents quantitatively using JIWC. ［Material］ (1) After recruiting 18 examinees of 53-90 years old (mean: 76.89), they were divided into two groups based on Mini Mental State Examination (MMSE) scores: the AD group (n＝9, MMSE≤21) and the healthy control (HC) group (n＝9, MMSE>21). (2) After recruiting 42 participants diagnosed dementia of 69-98 years old (mean: 19.79), they were divided into two groups based on Mini Mental State Examination (MMSE) scores: the Dementia group (n＝23, MMSE≤21) and mild cognitive impairment (MCI) group (n＝19, MMSE>21). ［Result］ Significant differences were confirmed for the usage of ＜anxiety＞ in the Dementia group and AD group. ［Discussion］ Generally, Dementia patients feel anxiety because of symptoms of Dementia. The results suggest that JIWC could quantitatively evaluate the Dementia features.

 Our team has been involved in developing a clinical decision support system (CDSS), which requires information about patients’ lifestyle. However, patients’ lifestyle issues are usually encoded in clinician generated narrative texts, which poses significant barriers to their information accessibility. In this paper, we propose an approach to identifying lifestyle issues of obesity, smoking and drinking in electronic health records (EHR) using machine learning and natural language processing techniques. To evaluate our approach, we conduct experiments using clinical narratives from The University of Tokyo Hospital which were generated in 2015 and saved in SS-MIX2 extended storage. The experimental results show that the proposed approach achieves equivalent high performance compared to previous studies focusing on English discharge summaries.