Published papers

Background Previous studies have attempted to characterize depression using electroencephalography (EEG), but results have been inconsistent. New noise reduction technology allows EEG acquisition during conversation. Methods We recorded EEG from 40 patients with depression as they engaged in conversation using a single-channel EEG device while conducting real-time noise reduction and compared them to those of 40 healthy subjects. Differences in EEG between patients and controls, as well as differences in patients’ depression severity, were examined using the ratio of the power spectrum at each frequency. In addition, the effects of medications were examined in a similar way. Results In comparing healthy controls and depression patients, significant power spectrum differences were observed at 3 Hz, 4 Hz, and 10 Hz and higher frequencies. In the patient group, differences in the power spectrum were observed between asymptomatic patients and healthy individuals, and between patients of each respective severity level and healthy individuals. In addition, significant differences were observed at multiple frequencies when comparing patients who did and did not take antidepressants, antipsychotics, and/or benzodiazepines. However, the power spectra still remained significantly different between non-medicated patients and healthy individuals. Limitations The small sample size may have caused Type II error. Patients’ demographic characteristics varied. Moreover, most patients were taking various medications, and cannot be compared to the non-medicated control group. Conclusion A study with a larger sample size should be conducted to gauge reproducibility, but the methods used in this study could be useful in clinical practice as a biomarker of depression.

Background: Although clozapine is effective for treatment-resistant schizophrenia (TRS), the rate of clozapine prescription is still low. Whereas antipsychotic monotherapy is recommended in clinical practice guidelines, the rate of antipsychotic polypharmacy is still high. There is little evidence on whether a clozapine prescription influences changes in the rate of monotherapy and polypharmacy, including antipsychotics and other psychotropics. We therefore hypothesized that the rate of antipsychotic monotherapy in patients with TRS who were prescribed clozapine would be higher than that in patients with schizophrenia who were not prescribed clozapine. Methods: We assessed 8306 patients with schizophrenia nationwide from 178 institutions in Japan from 2016 to 2019. We analyzed the psychotropic prescription data at discharge in patients diagnosed with TRS and with no description of TRS (ND-TRS) based on the diagnosis listed in the discharge summary. Results: The rate of antipsychotic monotherapy in the TRS with clozapine group (91.3%) was significantly higher than that in the TRS without clozapine group (45.9%; P< 2.0 × 10−16) and the ND-TRS without clozapine group (54.7%; P< 2.0 × 10−16). The rate of antipsychotic monotherapy without any other concomitant psychotropics in the TRS with clozapine group (26.5%) was significantly higher than that in the TRS without clozapine group (12.6%; P= 1.1 × 10−6) and the ND-TRS without clozapine group (17.0%; P= 5.9 × 10−6). Conclusions: Clozapine prescription could be associated with a high rate of antipsychotic monotherapy. Patients will benefit from the correct diagnosis of TRS and thus from proper clozapine prescription.

Background: Poor sleep quality can elevate stress levels and diminish overall well-being. Japanese individuals often experience sleep deprivation, and workers have high levels of stress. Nevertheless, research examining the connection between objective sleep assessments and stress levels, as well as overall well-being, among Japanese workers is lacking. Objective: This study aims to investigate the correlation between physiological data, including sleep duration and heart rate variability (HRV), objectively measured through wearable devices, and 3 states (sleepiness, mood, and energy) assessed through ecological momentary assessment (EMA) and use of rating scales for stress and well-being. Methods: A total of 40 office workers (female, 20/40, 50%; mean age 40.4 years, SD 11.8 years) participated in the study. Participants were asked to wear a wearable wristband device for 8 consecutive weeks. EMA regarding sleepiness, mood, and energy levels was conducted via email messages sent by participants 4 times daily, with each session spaced 3 hours apart. This assessment occurred on 8 designated days within the 8-week timeframe. Participants’ stress levels and perception of well-being were assessed using respective self-rating questionnaires. Subsequently, participants were categorized into quartiles based on their stress and well-being scores, and the sleep patterns and HRV indices recorded by the Fitbit Inspire 2 were compared among these groups. The Mann-Whitney U test was used to assess differences between the quartiles, with adjustments made for multiple comparisons using the Bonferroni correction. Furthermore, EMA results and the sleep and HRV indices were subjected to multilevel analysis for a comprehensive evaluation. Results: The EMA achieved a total response rate of 87.3%, while the Fitbit Inspire 2 wear rate reached 88.0%. When participants were grouped based on quartiles of well-being and stress-related scores, significant differences emerged. Specifically, individuals in the lowest stress quartile or highest subjective satisfaction quartile retired to bed earlier (P<.001 and P=.01, respectively), whereas those in the highest stress quartile exhibited greater variation in the midpoint of sleep (P<.001). A multilevel analysis unveiled notable relationships: intraindividual variability analysis indicated that higher energy levels were associated with lower deviation of heart rate during sleep on the preceding day (β=–.12, P<.001), and decreased sleepiness was observed on days following longer sleep durations (β=–.10, P<.001). Furthermore, interindividual variability analysis revealed that individuals with earlier midpoints of sleep tended to exhibit higher energy levels (β=–.26, P=.04). Conclusions: Increased sleep variabilities, characterized by unstable bedtime or midpoint of sleep, were correlated with elevated stress levels and diminished well-being. Conversely, improved sleep indices (eg, lower heart rate during sleep and earlier average bedtime) were associated with heightened daytime energy levels. Further research with a larger sample size using these methodologies, particularly focusing on specific phenomena such as social jet lag, has the potential to yield valuable insights.

Objective: To identify the COVID-19 pandemic impact on well-being/mental health, coping strategies, and risk factors in adolescents worldwide. Method: This study was based on an anonymous online multi-national/multi-language survey in the general population (representative/weighted non-representative samples, 14-17 years of age), measuring change in well-being (World Health Organization–Five Well-Being Index [WHO-5]/range = 0-100) and psychopathology (validated composite P-score/range = 0-100), WHO-5 <50 and <29, pre- vs during COVID-19 pandemic (April 26, 2020-June 26, 2022). Coping strategies and 9 a priori– defined individual/cumulative risk factors were measured. A χ2, penalized cubic splines, linear regression, and correlation analyses were conducted. Results: Analyzing 8,115 of 8,762 initiated surveys (representative = 75.1%), the pre-pandemic WHO-5 and P-score remained stable during the study (excluding relevant recall bias/drift), but worsened during the pandemic by 5.55 ± 17.13 (SD) and 6.74 ± 16.06 points, respectively (effect size d = 0.27 and d = 0.28). The proportion of adolescents with WHO-5 scores suggesting depression screening (<50) and major depression (<29) increased from 9% to 17% and 2% to 6%. WHO-5 worsened (descending magnitude, with cumulative effect) in adolescents with a mental or physical disorder, female gender, and with school closure. Results were similar for the P-score, with the exception of school closure (not significant) and living in a low-income country, as well as not living in a large city (significant). Changes were significantly but minimally related to COVID-19 deaths/restrictions, returning to near–pre-pandemic values after >2 years. The 3 most subjectively effective coping strategies were Internet use, exercise/walking, and social contacts. Conclusion: Overall, well-being/mental health worsened (small effect sizes) during the early stages of the COVID-19 pandemic, especially in vulnerable subpopulations. Identified at-risk groups, association with pandemic-related measures, and coping strategies can inform individual behaviors and global public health strategies. Plain language summary: The impact of COVID-19 and related restrictions on the mental health of children and adolescents remains unclear. This study conducted an anonymous online survey with over 8,100 adolescents aged 14 to 17 years worldwide. Results showed a significant decline in wellbeing, with depression rates increasing from 9% to 17% returning to near- pre-pandemic values after about 2 years. Adolescents with preexisting mental health conditions, female youth, and those affected by school closures were particularly vulnerable. Effective coping strategies included internet use, exercise, and social contacts, highlighting areas for public health focus to support youth during pandemics. Study preregistration information: Physical and mental health impact of COVID-19 on children, adolescents, and their families: The Collaborative Outcomes study on Health and Functioning during Infection Times - Children and Adolescents (COH-FIT-C&A); https://doi.org/10.1016/j.jad.2021.09.090 Diversity & Inclusion Statement: We worked to ensure sex and gender balance in the recruitment of human participants. We worked to ensure race, ethnic, and/or other types of diversity in the recruitment of human participants. We worked to ensure that the study questionnaires were prepared in an inclusive way. The author list of this paper includes contributors from the location and/or community where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work. We actively worked to promote inclusion of historically underrepresented racial and/or ethnic groups in science in our author group. We actively worked to promote sex and gender balance in our author group. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented racial and/or ethnic groups in science.

Pareidolia is a psychiatric phenomenon similar to visual illusions or hallucinations, serving as a distinguishing feature of psychiatric disorders. The so-called noise pareidolia test is a typical diagnosis method for pareidolia. In the test, ambiguous noise-like images are shown to patients, to see if patients misunderstand them as faces. However, the existing testing method is very time-consuming for both patients and experts. Furthermore, it has low accuracy in distinguishing some psychiatric disorders with similar characteristics. This research presents a machine learning-based framework to assist in assessing the pareidolia phenomenon. This is a challenging task because analyzing psychiatric disorders primarily involves observing behavioral changes. This is in contrast to other medical imaging tasks which rely on physically visible changes. Our framework emulates the behavior of patients in noise pareidolia tests. To achieve this, we fine-tune face detection models on patient data to misunderstand ambiguous patterns in the same way a patient would do. Then, we can compare new patients to the behavior of existing patients to assess the pareidolia phenomenon, through two classification approaches: a distance function-based approach and a multilayer perceptron-based approach. The proposed framework is evaluated on the data from 48 real patients. The result shows a performance superior to baseline methods in distinguishing pareidolia from other similar visual illusions. Moreover, our framework introduces two sampling methods to reduce the number of images shown to the patient while reaching a comparable high accuracy, improving the efficiency and making the framework feasible for a clinical test.

Background Clinical practice guidelines for schizophrenia and major depressive disorder have been published. However, these have not had sufficient penetration in clinical settings. We developed the Effectiveness of Guidelines for Dissemination and Education in Psychiatric Treatment (EGUIDE) project as a dissemination and education programme for psychiatrists. Aims The aim of this study is to assess the effectiveness of the EGUIDE project on the subjective clinical behaviour of psychiatrists in accordance with clinical practice guidelines before and 1 and 2 years after participation in the programmes. Method A total of 607 psychiatrists participated in this study during October 2016 and March 2019. They attended both 1-day educational programmes based on the clinical practice guidelines for schizophrenia and major depressive disorder, and answered web questionnaires about their clinical behaviours before and 1 and 2 years after attending the programmes. We evaluated the changes in clinical behaviours in accordance with the clinical practice guidelines between before and 2 years after the programme. Results All of the scores for clinical behaviours in accordance with clinical practice guidelines were significantly improved after 1 and 2 years compared with before attending the programmes. There were no significant changes in any of the scores between 1 and 2 years after attending. Conclusions All clinical behaviours in accordance with clinical practice guidelines improved after attending the EGUIDE programme, and were maintained for at least 2 years. The EGUIDE project could contribute to improved guideline-based clinical behaviour among psychiatrists.

There is no multi-country/multi-language study testing a-priori multivariable associations between non-modifiable/modifiable factors and validated wellbeing/multidimensional mental health outcomes before/during the COVID-19 pandemic. Moreover, studies during COVID-19 pandemic generally do not report on representative/weighted non-probability samples. The Collaborative Outcomes study on Health and Functioning during Infection Times (COH-FIT) is a multi-country/multi-language survey conducting multivariable/LASSO-regularized regression models and network analyses to identify modifiable/non-modifiable factors associated with wellbeing (WHO-5)/composite psychopathology (P-score) change. It enrolled general population-representative/weighted-non-probability samples (26/04/2020-19/06/2022). Participants included 121,066 adults (age=42±15.9 years, females=64 %, representative sample=29 %) WHO-5/P-score worsened (SMD=0.53/SMD=0.74), especially initially during the pandemic. We identified 15 modifiable/nine non-modifiable risk and 13 modifiable/three non-modifiable protective factors for WHO-5, 16 modifiable/11 non-modifiable risk and 10 modifiable/six non-modifiable protective factors for P-score. The 12 shared risk/protective factors with highest centrality (network-analysis) were, for non-modifiable factors, country income, ethnicity, age, gender, education, mental disorder history, COVID-19-related restrictions, urbanicity, physical disorder history, household room numbers and green space, and socioeconomic status. For modifiable factors, we identified medications, learning, internet, pet-ownership, working and religion as coping strategies, plus pre-pandemic levels of stress, fear, TV, social media or reading time, and COVID-19 information. In multivariable models, for WHO-5, additional non-modifiable factors with |B|>1 were income loss, COVID-19 deaths. For modifiable factors we identified pre-pandemic levels of social functioning, hobbies, frustration and loneliness, and social interactions as coping strategy. For P-scores, additional non-modifiable/modifiable factors were income loss, pre-pandemic infection fear, and social interactions as coping strategy. COH-FIT identified vulnerable sub-populations and actionable individual/environmental factors to protect well-being/mental health during crisis times. Results inform public health policies, and clinical practice.

Japan has seen an increase in female physicians recently, yet it still lags behind other Organization for Economic Cooperation and Development (OECD) countries. A major barrier has been the historical discrimination against women in medical school admissions. In recent years, female enrolment in medical schools has risen, surpassing 40% in 2024, reflecting a broader societal shift. However, structural problems persist in the Japanese healthcare system. Although the number of doctors per capita is lower in Japan than in other countries, the number of patients is higher than in other countries, leading to overwork for doctors. As a result, only about one-third of female doctors in Japan are able to return to work after interrupting their careers to give birth or raise children. The maldistribution of physicians, both regionally and by specialty, exacerbates this issue. To sustain the rising number of female physicians, Japan must reform its medical system.

Background: Mental health problems, particularly depression, significantly impact the working-age population and incur substantial socioeconomic costs. Electrodermal activity (EDA) and skin temperature, measurable through wearable devices, have emerged as potential biomarkers of depression. However, there is a lack of long-term observational studies in real-world settings. Methods: This 4-week observational study analyzed data from 147 office workers across 11 Japanese companies. Participants wore wearable devices during working hours for four weeks to collect EDA and skin temperature data. Depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9), and stress was assessed using the 10-item version of the Perceived Stress Scale (PSS-10). The associations between electrodermal activity, skin temperature, and depressive/stress states were analyzed using the Mann–Whitney U test and logistic regression analysis. Results: Participants with PHQ-9 scores ≥5 showed higher median skin conductance and median tonic signal, although these differences lost statistical significance after Bonferroni correction. Logistic regression analysis revealed a negative correlation between age and both PHQ-9 and PSS-10 scores (OR=0.525, 0.646). The mean and standard deviation of skin temperature showed a positive association with depressive symptoms (OR=1.687, 1.733), while normalized EDASymp showed a negative association (OR=0.564). Conclusions: This study suggests that EDA and skin temperature, measured by wearable devices, may be associated with depressive symptoms among office workers. These physiological indicators could be utilized for mental health monitoring and early intervention in the workplace. Further validation with larger samples, and long-term follow-up studies, are warranted.

© 2019 The Society of Instrument and Control Engineers - SICE. Depression is the most common mood disorder in the world which is also the leading cause of suicide. Depression might be diagnosed using varying modalities, as it affects many aspects of the patient, not only the patient will appear melancholic, but the patient's sleep quality and life quality might also be disrupted. In this study, we aim to extract facial landmark features unique to depression patients and build a machine learning model to classify the depression severity. The facial landmark activities considered in this study include: speed statistics of the landmarks, the standard deviation of eye pupils, and statistical features of mouth area. We found several facial landmark activity features with significant differences between healthy volunteers and depressed patients. We also successfully built machine learning models for automatic depression severity prediction.

© 2018 IEEE. Ballistocardiogram (BCG) is the repetitive body movements related to cardiac cycle. All night continuous measuring of BCG by using bed sensors enable us to detect sleep stage. However conventional methods are seemed to have unreliability caused by position of lying down. In this paper, we propose a reliable method of all night continuous measuring of BCG. Heart rate variability (HRV) is estimated from BCG measured with new bed sensors by template-matching approach. This approach has three steps, pre-processing, template-creation and template matching. Output of this process were validated with 4 subjects. The mean beat-to-beat interval error was 9.8[ms] and on an average 87.1% of the beat-to-beat intervals were detected.

© 2018 IEEE. Ballistocardiogram (BCG) is useful and informative signal result from the cardiac activity. In this paper, we present a practical BCG measurement system with bed sensors. Moreover, the simple algorithm to detect all individual heartbeats and estimate beat-to-beat interval lengths is presented. These are effective for all night continuous monitoring of cardiovascular system and enable us to detect sleep stage. Heart beats is detected by cross-correlation method. This method has three steps, preprocessing, typical wave form creation and cross-correlation calculation. Output of this process were validated with 10 subjects. As the results, the average 71.5% of the beat-to-beat intervals were detected. Moreover this study revealed that the individual difference is significant factor to detect heart beats precisely.

As the population of Japan continues to age, there is an increased need to conduct neuropsychological evaluations of seniors for conditions such as dementia.　Neuropsychological evaluations must be conducted by specialists who possess both strong technical abilities and a proficient understanding of higher brain functions, but as the number of patients requiring such evaluations continues to rise, there are not enough qualified specialists to meet the demand.　One way to correct this imbalance is to perform evaluations using telemedicine.　In parallel with this, the importance of using telemedicine has been further heightened by COVID-19 pandemic.　In this paper, we present:1） a review comparing the validity of in-person neuropsychological evaluations with that of evaluations conducted remotely using a video conferencing system （remote evaluations）;2） a review of patient satisfaction for in-person and remote evaluations;and 3） concrete processes that should be employed and important points to remember for remote evaluations, based on our team's experiences conducting them.　We found that studies reported a high reliability for remote evaluations even when performed on dementia patients.　Patient satisfaction with remote evaluations was generally favorable, but there were some complaints about difficulty hearing the reviewer via the remote system and the mental burden required to use the system/technology.　Regarding the processes required for remote evaluations, we recommend referring to the telepsychiatry manual created by the Telepsychiatry Guideline Manual Taskforce.　In the future, we hope that equitable access to guideline-adherent remote evaluations will become available across Japan.

The UNDERPIN large-scale mental disorder conversation corpus we have constructed consists of various diagnostic results of subjects and their conversation recordings, all of which have been transcribed manually. Furthermore, a portion of these have been annotated with detailed phonetic and linguistic annotations by hand. However, considering the practical use for future automated diagnostic support, automatic annotation is desirable. In our previous research, statistics on laughter as a vocal feature have been suggested to be an important clue for diagnosis regardless of the disorder, yet this too relies on manual annotation. In this study, we have annotated laughter in the UNDERPIN corpus and fine-tuned the speech recognition system Whisper to construct an automatic laughter detection system. The detection performance of laughter exceeded 90%, achieving a high level of practical performance. Using the constructed automatic laughter detection system, we annotated dialogue text data in the UNDERPIN corpus and conducted disease classification experiments. The results showed performance comparable to that using manual annotations, demonstrating a practical level of performance for applications.

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.