Published papers

This study aimed to clarify how behavioral and psychological symptoms of dementia (BPSD) and cognitive function affect the decision-making capacity of persons with Alzheimer's disease (AD) in a real informed consent situation about anti-dementia drug prescriptions. The participants were 76 patients with AD. We used the MacArthur Competence Assessment Tool to assess the capacity for consent to treatment (MacCAT-T). We simultaneously used the Mini-Mental State Examination, Executive Interview, Executive Clock Drawing Task, Logical Memory I of the Wechsler Memory Scale-Revised (LM I), LM II, and Neuropsychiatric Inventory (NPI) to assess cognitive function and psychiatric symptoms. We calculated the correlations between the MacCAT-T scores and the demographic, neuropsychological, and psychiatric variables. Once the univariable correlations were determined, we performed simple linear regression analyses to examine if the regression equations were significant. In the final analyses, we incorporated significant variables into stepwise multiple linear regression analyses to determine the most significant predictors of mental capacity. Age (β = −0.34), anxiety (β = −0.27), and LM I (β = 0.26) were significant predictors of “understanding” (adjusted R2 = 0.29). LM II (β = 0.39), anxiety (β = −0.29), and education (β = 0.21) were significant predictors of “understanding of alternative treatments” (adjusted R2 = 0.30). Anxiety (β = −0.36) and age (β = −0.22) were significant predictors of “appreciation” (adjusted R2 = 0.18). Age (β = −0.31) and anxiety (β = −0.28) were significant predictors of explained variance in “reasoning” (adjusted R2 = 0.17). Patients with anxiety had lower scores on all five MacCAT-T subscales: “understanding,” without 3.8 [SD = 1.2] vs. with 2.6 [SD = 1.1]; “understanding of alternative treatments,” without 2.9 [SD = 2.2] vs. with 1.3 [SD = 1.8]; “appreciation,” without 2.9 [SD = 1.1] vs. with 1.9 [SD = 1.2]; “reasoning,” without 4.0 [SD = 2.0] vs. with 2.7 [SD = 1.7]; and “expressing a choice,” without 1.9 [SD = 0.4] vs. with 1.5 [SD = 0.6]. Considering the effects of BPSD, cognitive function, and age/education when assessing consent capacity in persons with AD is important. Reducing anxiety may contribute to improved capacity in persons with AD.

Artificial intelligence (AI) can provide many benefits in healthcare, including rapid and effective treatment options. However, previous research on human–computer interactions has demonstrated that people are reluctant to accept AI. This study compared individuals’ trust in AI with their trust in doctors regarding medical treatment and explored whether people trust an AI system that understands and suggests their desired treatments. We conducted an online experiment (N = 415) adopting a scenario method. The participants were asked to imagine receiving a medical examination and being prescribed medicine by a physician or an AI system. Participants trusted the AI system less than a doctor, even when the AI system learned and suggested their desired treatment. This finding confirms that people are reluctant to trust AI even if it performs at the level of a human doctor, thereby strengthening the existent literature that shows people prefer human doctors.

Some older adults desire to continue living in the community in which they reside. The concept of Aging in Place (AIP) is appropriate for this idea. For AIP to be successful, the oldest-old need to demonstrate their abilities and live safely, independently, and comfortably in their homes and communities. Conversely, AIP in an inappropriate manner, without a safe environment, and without support provided, comprising dangerous situations caused by declining cognitive and physical function, lead to reduced quality of life for older people. Therefore, this study aimed to examine cognitive function in the oldest-old to understand differences between normal aging and cognitive decline due to dementia. Cognitive decline due to normal aging in the oldest-old is caused by impaired frontal lobe function, while decline due to Alzheimer’s disease is caused by impaired temporal-parietal lobe function, suggesting that this impairment may be useful for achieving independence for AIP.

Objective: To develop and validate the Testamentary Capacity Assessment Screening Tool (TCAST) for older adults, addressing the absence of standardised instruments for testamentary capacity (TC) assessment in Japan. Methods: The TCAST was developed through an iterative process including literature review, interdisciplinary expert consultation, and user testing with older adults. It comprises three subtests: (1) Hypothetical will-making Scenario, (2) Basic financial concepts, and (3) Legal vocabulary. Internal consistency, test–retest reliability, and construct validity were examined using Cronbach's α, intraclass correlation coefficients (ICC), and correlations with the Hasegawa Dementia Scale-Revised (HDS-R), the Executive Interview (EXIT25), and the Geriatric Depression Scale (GDS-15). Group comparisons and receiver operating characteristic (ROC) analyses were conducted to evaluate discriminatory performance. Results: A total of 118 participants—68 cognitively healthy older adults and 50 with dementia—completed the TCAST. Internal consistency was good for Parts 1 and 2 (α = 0.88, 0.70) and lower for Part 3 (α = 0.58). Test–retest reliability was excellent for Parts 1 and 2 (ICC = 0.87, 0.89) and moderate for Part 3 (ICC = 0.58). Construct validity was supported by moderate to strong correlations with HDS-R and EXIT25, but not with GDS-15. All subtests correlated negatively with age and positively with education, and significantly differentiated healthy and dementia groups (p < 0.001). ROC analyses showed strong discriminatory capacity (AUC = 0.86, 0.97, and 0.79 for Parts 1–3). Conclusions: The TCAST is a reliable and valid screening tool for assessing TC in older Japanese adults. It is not designed to categorically determine incapacity, but to highlight domains relevant to testamentary reasoning and support person-centered evaluations. The tool may aid clinicians and legal professionals in identifying both retained abilities and areas needing support, underscoring its potential utility in interdisciplinary approaches to capacity assessment in aging societies.

Aim: Telepsychiatry is increasingly integral to psychiatric practice. However, few reviews have examined the concordance between assessments conducted in telepsychiatric and face-to-face settings. This systematic review and meta-analysis evaluated the agreement between telepsychiatric and face-to-face settings in the diagnosis and symptom assessment of various psychiatric disorders. Methods: A literature search was conducted using MEDLINE/PubMed, Cochrane Library, Scopus, EMBASE, CINAHL, and PsycINFO. Studies evaluating the concordance between telepsychiatric and face-to-face settings in the diagnosis and symptom assessment of various psychiatric disorders were included and analyzed. Results: Of the 6875 studies in the initial search, 22 studies that met the inclusion criteria were included in the analyses. The diagnostic concordance for 16 psychiatric disorders was “almost perfect” between the two settings (N = 16, n = 848, Cohen's κ = 0.824, confidence interval [CI] = 0.466 to 0.950, P < 0.001). Additionally, the concordance for the symptom rating scales between the two settings ranged from “substantial” in the Brief Psychiatric Rating Scale (N = 1, n = 533, Cohen's κ = 0.789, CI = 0.699 to 0.855, P < 0.001) to “almost perfect” in the Autism Diagnostic Observation Schedule (N = 1, n = 92, intraclass correlation coefficients = 0.943, CI = 0.798 to 0.985, P < 0.001). Conclusions: Telepsychiatry is highly concordant with face-to-face settings regarding psychiatric diagnoses and symptom assessment. In contrast, the present results also suggest that the suitability of telepsychiatry varies across disease types, specific symptoms, and assessment modalities. Although the present results must be interpreted with caution owing to the small number of studies for each assessment and disease, our findings suggest that telepsychiatry may have greater utility in psychiatric diagnostic assessment settings.

INTRODUCTION: Clinical features of cognitive performance in extreme old age differ from those of pathological cognitive decline in Alzheimer's disease (AD). METHODS: We compared cognitive traits between 638 centenarians aged 100–115 years and 208 and 221 patients with AD from independent facilities. RESULTS: The presence of the apolipoprotein E (APOE) ε4 allele did not affect Mini-Mental State Examination (MMSE) scores in centenarians. Centenarians retained the ability to follow three consecutive commands, associated with their educational background and activities of daily living. Cognitive retention remained unchanged in semi-supercentenarians (aged ≥ 105 years) and supercentenarians (aged ≥110 years). A quantitative genome-wide association study (GWAS) identified two loci associated with maintaining the ability to follow three consecutive commands. DISCUSSION: This is the first study to compare cognitive traits between >600 centenarians and patients with AD. Centenarians attained higher MMSE scores for the phenotype of following three consecutive commands than patients with AD, being useful in clinical practice. Highlights: Cognitive phenotypes in centenarians differ from those in the AD groups Clinical trait to follow consecutive instructions is retained in centenarians but not in AD groups GWAS identified SNPs related to the maintained trait of MMSE in centenarians.

Objectives: Treatment decisions for severe aortic stenosis (AS) are complex, since there are two active and comparable options: transcatheter aortic valve implantation and surgical aortic valve replacement. The disease predominantly affects older individuals, who are frequently comorbid with from cognitive impairment. This study aimed to establish a screening-triggered system to assess the decision-making capacity of patients with AS, support their decision-making, and facilitate referrals to specialists when necessary. Methods: This is a retrospective pre-post observational study. Relevant healthcare professionals were trained to assess and support patients' decision-making capacities. Results: Subtotals of 203 and 244 patients were enrolled before and after the implementation of the system, respectively. The requests for decision-making support significantly increased (from 1.5 % to 14.8 % of the patients). 11.5 % of the patients were identified as with declined decision-making capacity. The proportion of patients who received active treatments for AS significantly decreased from 95.0 % to 83.0 %. The nurses' documentation of patients' decision-making capacities significantly increased. Conclusion: This screening-triggered system identified a substantial proportion of patients with declined decision-making capacity. A significant decrease in patients receiving active treatments for AS and increased documentation of patients' decision-making capacities in nursing records were observed. A future randomized controlled is warranted.

In recent years, studies on the use of natural language processing (NLP) approaches to identify dementia have been reported. Most of these studies used picture description tasks or other similar tasks to encourage spontaneous speech, but the use of free conversation without requiring a task might be easier to perform in a clinical setting. Moreover, free conversation is unlikely to induce a learning effect. Therefore, the purpose of this study was to develop a machine learning model to discriminate subjects with and without dementia by extracting features from unstructured free conversation data using NLP. We recruited patients who visited a specialized outpatient clinic for dementia and healthy volunteers. Participants’ conversation was transcribed and the text data was decomposed from natural sentences into morphemes by performing a morphological analysis using NLP, and then converted into real-valued vectors that were used as features for machine learning. A total of 432 datasets were used, and the resulting machine learning model classified the data for dementia and non-dementia subjects with an accuracy of 0.900, sensitivity of 0.881, and a specificity of 0.916. Using sentence vector information, it was possible to develop a machine-learning algorithm capable of discriminating dementia from non-dementia subjects with a high accuracy based on free conversation.

Aim: The authors applied natural language processing and machine learning to explore the disease-related language patterns that warrant objective measures for assessing language ability in Japanese patients with Alzheimer disease (AD), while most previous studies have used large publicly available data sets in Euro-American languages. Methods: The authors obtained 276 speech samples from 42 patients with AD and 52 healthy controls, aged 50 years or older. A natural language processing library for Python was used, spaCy, with an add-on library, GiNZA, which is a Japanese parser based on Universal Dependencies designed to facilitate multilingual parser development. The authors used eXtreme Gradient Boosting for our classification algorithm. Each unit of part-of-speech and dependency was tagged and counted to create features such as tag-frequency and tag-to-tag transition-frequency. Each feature's importance was computed during the 100-fold repeated random subsampling validation and averaged. Results: The model resulted in an accuracy of 0.84 (SD = 0.06), and an area under the curve of 0.90 (SD = 0.03). Among the features that were important for such predictions, seven of the top 10 features were related to part-of-speech, while the remaining three were related to dependency. A box plot analysis demonstrated that the appearance rates of content words–related features were lower among the patients, whereas those with stagnation-related features were higher. Conclusion: The current study demonstrated a promising level of accuracy for predicting AD and found the language patterns corresponding to the type of lexical-semantic decline known as ‘empty speech’, which is regarded as a characteristic of AD.

Research on elderly financial exploitation has mostly focused on financial abuse that occurs in families and other types of trusted relationships. As such, little is known about financial frauds and scams perpetrated by strangers. Financial fraud and scam prevention activities for older adults must be promoted, for which the correlation between the psychological, social, and cognitive characteristics of their vulnerability needs to be determined. The present study aimed to determine whether cognitive decline is a risk factor for scam vulnerability in older adults. Thus, we created a scam vulnerability scale for older adults with cognitive decline and analyzed the data to reveal the correlation between them, including inhibition and executive function. We conducted an interview survey with 50 older adults with cognitive decline (average age: 79.42 years, SD: 5.44) and 51 older adults without cognitive decline (average age: 76.12 years, SD: 5.82). The interview survey included the scam vulnerability scale, psychosocial questionnaires, and neuropsychological tests. The scale included six items with a four-point Likert scale based on a previous study. Hierarchical multiple regression analysis revealed that lower scores on the Japanese version of the Alzheimer's Disease Assessment Scale–Cognitive Subscale (ADAS-Jcog; indicating higher general cognitive function) correlated with higher scam vulnerability in the cognitive decline group (β = −0.46, p < 0.001). In addition, men were found to be more vulnerable in both groups (cognitive decline group: β = −0.29, p = 0.015, cognitive non-decline group; β = −0.32, p = 0.018). Inhibition and executive function were found not to correlate significantly with scam vulnerability. These results suggest that mild cognitive decline correlates with higher scam vulnerability, whereas moderate to severe cognitive decline correlates with lower vulnerability, possibly because it makes understanding the scam attempt itself difficult. Older adults with mild cognitive decline and their families, particularly those visiting elderly care or outpatient facilities, should be notified of the scam vulnerability of older clients using the ADAS-Jcog score as an index to help them avoid victimization.

Objective: To clarify the relationship between life satisfaction and the psychological characteristics of the oldest-old, and explore the factors for achieving mental health and longevity. Design: This cross-sectional study conducted questionnaire surveys and face-to-face interviews as part of a larger prospective cohort study. Setting: Arakawa Ward, a district in Tokyo, Japan. Participants: A total of 247 oldest-old individuals from two age groups, 85+ (aged 85–87 years) and 95+ (aged 95 years or older). Measurements: Life satisfaction was assessed using the Satisfaction with Life Scale (SWLS), developmental stages of the elderly (Erikson's 8th and 9th stages, i.e., ego integrity, and gerotranscendence), and the Big Five personality traits (extraversion, agreeableness, openness, conscientiousness, and neuroticism) using the NEO-Five Factor Inventory. Multiple regression analyses were performed to examine the relationship between the SWLS scores and each assessment, controlling for age, sex, education, activities of daily living, depressive symptoms, and cognitive function. Results: The SWLS scores of 85+ were positively correlated with scores of ego integrity, extraversion, and conscientiousness. Contrastingly, the SWLS scores of 95+ were positively correlated with gerotranscendence scores. Conclusions: Psychological characteristics associated with the level of life satisfaction among community-dwelling oldest-old individuals were identified, but a causal relationship between these factors and life satisfaction was not established. Ego integrity, extraversion, conscientiousness, and gerotranscendence may be associated with enhanced life satisfaction and mental health in the oldest-old. Further, the factors associated with life satisfaction in the 85+ and 95+ age groups varied, suggesting that life satisfaction among the oldest-old has different foundations in different age groups.

We propose a non-invasive and cost-effective method to automatically detect dementia by utilizing solely speech audio data. We extract paralinguistic features for a short speech segment and use Gated Convolutional Neural Networks (GCNN) to classify it into dementia or healthy. We evaluate our method on the Pitt Corpus and on our own dataset, the PROMPT Database. Our method yields the accuracy of 73.1% on the Pitt Corpus using an average of 114 seconds of speech data. In the PROMPT Database, our method yields the accuracy of 74.7% using 4 seconds of speech data and it improves to 80.8% when we use all the patient’s speech data. Furthermore, we evaluate our method on a three-class classification problem in which we included the Mild Cognitive Impairment (MCI) class and achieved the accuracy of 60.6% with 40 seconds of speech data.

Background The COVID-19 pandemic has transformed healthcare significantly and telepsychiatry is now the primary means of treatment in some countries. Aims To compare the efficacy of telepsychiatry and face-to-face treatment. Method A comprehensive meta-analysis comparing telepsychiatry with face-to-face treatment for psychiatric disorders. The primary outcome was the mean change in the standard symptom scale scores used for each psychiatric disorder. Secondary outcomes included all meta-analysable outcomes, such as all-cause discontinuation and safety/tolerability. Results We identified 32 studies (n = 3592 participants) across 11 mental illnesses. Disease-specific analyses showed that telepsychiatry was superior to face-to-face treatment regarding symptom improvement for depressive disorders (k = 6 studies, n = 561; standardised mean difference s.m.d. = -0.325, 95% CI -0.640 to -0.011, P = 0.043), whereas face-to-face treatment was superior to telepsychiatry for eating disorder (k = 1, n = 128; s.m.d. = 0.368, 95% CI 0.018-0.717, P = 0.039). No significant difference was seen between telepsychiatry and face-to-face treatment when all the studies/diagnoses were combined (k = 26, n = 2290; P = 0.248). Telepsychiatry had significantly fewer all-cause discontinuations than face-to-face treatment for mild cognitive impairment (k = 1, n = 61; risk ratio RR = 0.552, 95% CI 0.312-0.975, P = 0.040), whereas the opposite was seen for substance misuse (k = 1, n = 85; RR = 37.41, 95% CI 2.356-594.1, P = 0.010). No significant difference regarding all-cause discontinuation was seen between telepsychiatry and face-to-face treatment when all the studies/diagnoses were combined (k = 27, n = 3341; P = 0.564). Conclusions Telepsychiatry achieved a symptom improvement effect for various psychiatric disorders similar to that of face-to-face treatment. However, some superiorities/inferiorities were seen across a few specific psychiatric disorders, suggesting that its efficacy may vary according to disease type.

Introduction: Psychiatric disorders are diagnosed through observations of psychiatrists according to diagnostic criteria such as the DSM-5. Such observations, however, are mainly based on each psychiatrist's level of experience and often lack objectivity, potentially leading to disagreements among psychiatrists. In contrast, specific linguistic features can be observed in some psychiatric disorders, such as a loosening of associations in schizophrenia. Some studies explored biomarkers, but biomarkers have yet to be used in clinical practice. Aim: The purposes of this study are to create a large dataset of Japanese speech data labeled with detailed information on psychiatric disorders and neurocognitive disorders to quantify the linguistic features of those disorders using natural language processing and, finally, to develop objective and easy-to-use biomarkers for diagnosing and assessing the severity of them. Methods: This study will have a multi-center prospective design. The DSM-5 or ICD-11 criteria for major depressive disorder, bipolar disorder, schizophrenia, and anxiety disorder and for major and minor neurocognitive disorders will be regarded as the inclusion criteria for the psychiatric disorder samples. For the healthy subjects, the absence of a history of psychiatric disorders will be confirmed using the Mini-International Neuropsychiatric Interview (M.I.N.I.). The absence of current cognitive decline will be confirmed using the Mini-Mental State Examination (MMSE). A psychiatrist or psychologist will conduct 30-to-60-min interviews with each participant; these interviews will include free conversation, picture-description task, and story-telling task, all of which will be recorded using a microphone headset. In addition, the severity of disorders will be assessed using clinical rating scales. Data will be collected from each participant at least twice during the study period and up to a maximum of five times at an interval of at least one month. Discussion: This study is unique in its large sample size and the novelty of its method, and has potential for applications in many fields. We have some challenges regarding inter-rater reliability and the linguistic peculiarities of Japanese. As of September 2022, we have collected a total of >1000 records from >400 participants. To the best of our knowledge, this data sample is one of the largest in this field. Clinical Trial Registration: Identifier: UMIN000032141.

© 2020 Japan Geriatrics Society Aim: To investigate the influence of replacing sedentary time with physical activity on cognitive function using an isotemporal substitution model in a population of community-dwelling oldest old. Methods: This cross-sectional study included residents of the Arakawa ward, Tokyo, who were part of a prospective cohort from the Arakawa 85+ study. We measured physical activity in 136 participants using a triaxial actigraph. Cognitive function was measured using the Addenbrooke's Cognitive Examination-III and participants were divided into a “cognitive decline group” (Addenbrooke's Cognitive Examination-III ≤88) and “cognitive maintain group” (Addenbrooke's Cognitive Examination-III ≥89). Physical activity was divided into three categories: sedentary behavior (≤1.5 metabolic equivalents), light physical activity (>1.5 to <3.0 metabolic equivalents), and moderate-to-vigorous physical activity (≥3 metabolic equivalents). Using an isotemporal substitution approach, we applied multiple logistic regression analysis to demonstrate the association between cognitive function and replacing 30 min/day of sedentary behavior with an equal period of light physical activity. Covariates included age, education and the Center for Epidemiologic Studies Depression Scale. Results: Our findings showed that in men, replacing 30 min of sedentary behavior per day with light physical activity was associated with a 1.47-fold increase in the odds of maintaining cognitive function. An association between physical activity and cognitive function was not observed in female participants. Conclusions: Our results indicate that substituting sedentary behavior with light physical activity could be helpful in maintaining cognitive function in community-dwelling oldest old men. These results highlight the importance of behavioral changes to promote cognition. Geriatr Gerontol Int ••; ••: ••–•• Geriatr Gerontol Int 2020; ••: ••–••.

© 2018, © 2018 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Purpose: Current healthcare systems are not suitable for serving future societies in which the oldest old are commonplace. The objective of this study was to understand what the oldest old care most about in their daily lives. Methods: Semi-structured in-depth interviews and thematic analysis were used. Face-to-face interviews were conducted in 17 elderly residents (≥ 95 years) of Arakawa-ku, Tokyo, Japan from July to November 2017. Results: Three themes emerged from the interview responses: “unshakable beliefs and social ties,” “natural acceptance,” and “my day-to-day life with precious moments.” The oldest old strongly believe in diligence and compassion and maintain strong relationships with people around them. Despite their small social networks, they are concerned about future society. They accept their selves and their lives, including their impending deaths. Despite their functional decline, they control their lives by making very small decisions. They live on a moment-to-moment basis, cherishing simple events. Conclusion: Maintaining autonomy through making small decisions and enjoying small pleasures are important to the oldest old. Understanding the needs of the oldest old is the first step towards developing optimal geriatric care for an aging population.

© 2018 Komeda, Eguchi, Kusumi, Kato, Narumoto and Mimura. Information used by older adults engaging in a social decision making task of judging a protagonist as a good or a bad person was investigated. Older (n = 100, 50 women, mean age = 63.6 years) and younger (n = 100, 50 women, mean age = 25.7 years) adults participated in a web-based survey. In Experiment 1, we assessed participants' rapid decision-making processes when making good or bad judgments after reading consecutive sentences without reviewing previously read sentences. The percentages of good judgments were analyzed. In Experiment 2, two protagonists engaging in a deliberate decision-making process were presented, and participants were asked to judge better and worse protagonists. The percentages of behavior-based judgments were analyzed. Results of Experiment 1 indicated that older adults judged protagonists as "good" more often than younger adults. Especially, older adults judged protagonists with good behavior as being "good." In Experiment 2, older adults made behavior-based judgments more than younger people. Additionally, older and younger adults used information on personalities of protagonists for making judgments in situations with bad outcomes, or incongruent. Moreover, multiple regression analysis suggested that people with more general trust engaged more, whereas people with more caution engaged less in making behavior-based judgments.

Copyright © 2018 John Wiley & Sons, Ltd. Background: Despite a steady increase in life expectancy, a few studies have investigated cross-sectional correlates and longitudinal predictors of cognitive function, a core domain of the successful aging, among socio-clinico-demographic factors in the oldest-old exclusively. Objectives: The aims of this study were to examine socio-clinico-demographic characteristics associated with global cognition and its changes in the oldest-old. Methods: We reanalyzed a dataset of cognitively preserved community-dwelling subjects aged 85 years and older in the Tokyo Oldest Old Survey on Total Health, a 6-year longitudinal observational study. This study consisted of (1) baseline cross-sectional analyses examining correlates of global cognition (n = 248) among socio-clinico-demographic factors and (2) longitudinal analyses examining baseline predictors for changes of global cognition in 3-year follow-up (n = 195). The Mini-Mental State Examination was used as a screening test to assess global cognition. Results: At baseline, higher weights were related to higher cognitive function in the oldest-old. The baseline predictors of global cognitive decline in 3-year follow-up were higher global cognition, shorter education period, and lower sociocultural activities and lower instrumental activity of daily living, in this order. Conclusions: The present study suggests that it is crucial to attain higher education during early life and avoid leanness or obesity, participate in sociocultural cognitive activities during late life, and maintain instrumental activity of daily living to preserve optimal cognitive function in the oldest-old, which will facilitate developing prevention strategies for cognitive decline and promoting successful aging in this increasing population.

© Springer Nature Switzerland AG 2020. Recently, the number of people who are diagnosed as mental diseases is increasing. Efficient and objective diagnosis is important to start medical treatments in earlier stages. However, mental disease diagnosis is difficult to quantify criteria, because it is performed through conversations with patients, not by physical surveys. We aim to automate mental disease diagnosis in order to resolve these issues. We recorded conversations between psychologists and subjects to build our diagnosis speech corpus. Our subjects include healthy persons, people with mental diseases of depression, bipolar disorder, schizophrenia, anxiety and dementia. All of our subjects are diagnosed by doctors of psychiatry. Then we made accurate transcription manually, adding utterance time stamps, linguistic and non-linguistic annotations. Using our corpus, we performed feature analysis to find characteristics for each disease. We also tried automatic mental disease diagnosis by machine learning, while the number of sample data is few because we were still in our pilot study phase. We will increase the number of subjects in future.

© The Author(s) 2019. Introduction: Studies have demonstrated the high agreement of several remote neuropsychological tests using video teleconferencing (VTC) with face-to-face (FTF) tests. However, the reliability of the remotely administered Alzheimer’s Disease Assessment Scale cognitive subscale (ADAS-cog), one of the most commonly used neuropsychological tests to detect cognitive decline, has not been substantially elucidated, particularly in Japanese populations. Therefore, this study aimed to evaluate the reliability of the remotely administered ADAS-cog compared with FTF-administered ADAS-cog among elderly Japanese participants. Methods: Participants aged ≥60 years with and without cognitive impairment, i.e. those with mild cognitive impairment (MCI), those with dementia and healthy controls (HCs), were assessed with the ADAS-cog using VTC and FTF testing at an interval of >2 weeks and <3 months. The assessment order (VTC or FTF) was randomized by participants. Participants’ scores were compared among the entire sample, as well as subgroups, using intra-class correlation coefficients (ICCs) in a mixed-effects model. Results: A total of 73 participants were included in the study (36 men; age, 76.3 ± 7.6 years). The ICC for the ADAS-cog total score was high in the entire sample (0.86), whereas ICCs were moderate to high for the subgroups (MCI: 0.63, dementia: 0.80 and HC: 0.74). Discussion: The results indicate that a VTC-administered ADAS-cog could be an alternative for an FTF-administered ADAS-cog, although further replication studies with larger sample sizes and a wider range of cognitive functionalities are warranted.