As the global burden of dementia continues to plague our healthcare system, efficient, objective, and sensitive tools to characterize cognition and detect underlying neurodegenerative disease increasingly are needed. Digital phenotyping relies on passive, continuous collection of smartphone sensor data during everyday life to measure activities, behaviors, and mood. The present study explored the feasibility, acceptability, and validity of a digital phenotyping protocol as a novel method for characterizing cognition and function among a heterogeneous group of older adults. Validation analyses were based on a recently proposed conceptual model explaining activity level and variability as a function of cognitive ability level. Exploratory analyses aimed to examine and account for a range of participant and environmental factors that may be associated with digital phenotyping data. A total of 22 participants ages 65 - 81 years with either healthy cognition or mild cognitive impairment (MCI) used their own personal smartphones naturally during a four-week study period while a secure software application unobtrusively and continuously obtained Global Positioning System (GPS)-based movement trajectories. Participants completed gold-standard neuropsychological measures and questionnaires of everyday function, mood, and mobility habits at a baseline visit intended to evaluate construct validity. In-depth informed consent and a comprehension of consent quiz also were administered at baseline to inform feasibility of explaining digital phenotyping study procedures to older adults. Debriefing questionnaires were completed at the end of the study period, including questions pertaining to acceptability. Correlation analyses showed that measures of GPS activity and variability were positively associated with validators of cognition, everyday function, mood, and mobility habits. Potential confounding factors included season of study participation, unexpected health changes, and highest lifetime household annual income, whereas participant demographics such as education, sex, and race were not significantly associated with GPS features. Metrics on study withdrawal, comprehension of consent, and satisfaction ratings at study completion revealed good feasibility and acceptability. In sum, digital phenotyping shows promise as a feasible, acceptable, and potentially valid method to efficiently and objectively assess cognition, function, and mood in a cohort of older adults. Future studies will benefit from incorporating these preliminary findings and testing predictions in larger, more diverse cohorts both cross-sectionally and over time in longitudinal designs. / Psychology

BACKGROUND: As smartphone devices have become a ubiquitous part of our modern lives, parents and clinicians have become increasingly concerned about the effects of digital media use on the mental well-being of adolescents and young adults. Smartphone ownership in youth has increased significantly over the last decade, paralleling the rise in mental health disorders. This study seeks to use the digital phenotyping (DP) methodology to elucidate these relationships. Most studies examining these variables use cross-sectional data in healthy adolescents. To our knowledge, no studies have used DP methodology to characterize the relationship between digital media use, depression and anxiety in a population of clinical adolescents. METHODS: 50 adolescent and young adults between the ages of 12-23 receiving outpatient mental health services from a community hospital network in the greater Boston area were enrolled. Participants installed an application on their personal smartphones that collected daily surveys that captured mood symptoms, digital media use (screen time, social media time, and top apps used [active data]), and that also continuously captured sensor data (GPS and accelerometer [passive data]) over six weeks. RESULTS: Using linear regression and multilevel modeling, no significant associations were found between screen time or social media time, and anxiety and depression symptoms. Productivity apps were used significantly more in those with no depression symptoms than in those with moderate to severe levels of depression. CONCLUSION: Our study results challenge the present intuition that the amount of digital media use negatively impacts mental well-being in youth. Total screen time and social media time measures may be insufficient when attempting to assess the impact of digital media engagement on youth. Additionally, the results of our study suggest that the types of apps used by youth may depend on an individual’s mood severity. Although not without limitations, DP studies may be the ideal methodology for capturing with greater granularity digital use behavior and its association with mood symptoms in adolescents. / 2022-11-22T00:00:00Z

This dissertation investigates the utility of deep learning and machine learning approaches for livestock management and quantitative genetic modeling of rice grain size under climate change. Monitoring the live body weight of animals is crucial to support farm management decisions due to its direct relationship with animal growth, nutritional status, and health. However, conventional manual weighing methods are time consuming and can cause potential stress to animals. While there is a growing trend towards the use of three-dimensional cameras coupled with computer vision techniques to predict animal body weight, their validation with deep learning models as well as large-scale data collected in commercial environments is still limited. Therefore, the first two research chapters show how deep learning-based computer vision systems can enable accurate live body weight prediction for dairy cattle and pigs. These studies also address the challenges of managing large, complex phenotypic data and highlight the potential of deep learning models to automate data processing and improve prediction accuracy in an industry-scale commercial setting. The dissertation then shifts the focus to crop resilience, particularly in rice, where the asymmetric increase in average nighttime temperatures relative to the increase in average daytime temperatures due to climate change is reducing grain yield and quality in rice. Through the use of deep learning and machine learning models, the last two chapters explore how metabolic data can be used in quantitative genetic modeling in rice under environmental stress conditions such as high night temperatures. These studies showed that the integration of metabolites and genomics provided an improvement in the prediction of rice grain size-related traits, and certain metabolites were identified as potential candidates for improving multi-trait genomic prediction. Further research showed that metabolic accumulation was low to moderately heritable, and genomic prediction accuracies were consistent with expected genomic heritability estimates. Genomic correlations between control and high night temperature conditions indicated genotype-by-environment interactions in metabolic accumulation and the effectiveness of genomic prediction models for metabolic accumulation varied across metabolites. Joint analysis of multiple metabolites improved the accuracy of genomic prediction by exploiting correlations between metabolite accumulation. Overall, this dissertation highlights the potential of integrating digital technologies and multi-omic data to advance data analytics in agriculture, with applications in livestock management and quantitative genetic modeling of rice. / Doctor of Philosophy / This dissertation explores the application of deep learning and machine learning to computer vision-based livestock management and quantitative genetic modeling of rice grain size under climate change. The first half of the research chapters illustrate how computer vision systems can enable digital phenotyping of dairy cows and pigs, which is critical for informed management decisions and quantitative genetic analysis. These studies address the challenges of managing large-scale, complex phenotypic data and highlight the potential of deep learning models to automate data processing and improve prediction accuracy. Chapter 3 showed that a deep learning-based segmentation, Mask R-CNN, improved the prediction performance of cow body weight from longitudinal depth video data. Among the image features, volume followed by width correlated best with body weight. Chapter 4 showed that efficient deep learning-based supervised learning models are a promising approach for predicting pig body weight from industry-scale depth video data. Although the sparse design, which simulates budget and time constraints by using a subset of the data for training, resulted in some performance loss compared to the full design, the Vision Transformer models effectively mitigated this loss. The second half of the research chapters focuses on integrating metabolomic and genomic data to predict grain traits and metabolic content in rice under climate change. Through the use of machine learning models, these studies investigate how combining genomic and metabolic data can improve predictions, particularly under high night temperature stress in rice. Chapter 5 showed that the integration of metabolites and genomics improved the prediction of rice grain size-related traits, and certain metabolites were identified as potential candidates for improving multi-trait genomic prediction. Chapter 6 showed that metabolic accumulation was low to moderately heritable. Genomic correlations between control and high night temperature conditions indicated genotype-by-environment interactions in metabolic accumulation, and the effectiveness of genomic prediction models for metabolic accumulation varied across metabolites. Joint analysis of multiple metabolites improved the accuracy of genomic prediction by exploiting correlations between metabolite accumulation. Overall, the dissertation provides insight into how cutting-edge methods can be used to improve livestock management and multi-omic quantitative genetic modeling for breeding.

[pt] O presente trabalho aborda o monitoramento remoto em tempo real de pessoas com Transtorno Afetivo Bipolar e sua interação com seus dispositivos móveis. Algumas abordagens na área da ciência da computação apresentam experiências para coletar informações subjetivas as quais podem ser usadas por especialistas em benefício de pessoas com necessidades específicas. Por outra parte, a fenotipagem digital é um campo de ciência multidisciplinar, usado para descrever uma nova abordagem para acompanhar a interação dos usuários com seu smartphone e poder mapear transtornos fazendo uso dos sensores do celular. Embora algumas pesquisas avaliem as vantagens dos aplicativos móveis para o tratamento das doenças mentais, na literatura não foi encontrada nenhuma solução que faça uso da análise do fenótipo digital e a visualização da informação em tempo real como marcadores de estado e traço de intervenção terapêutica não invasiva, ao detectar precocemente alterações nos padrões comportamentais dos pacientes. Neste trabalho, apresentamos o BraPolar, uma m-Health para monitoramento remoto de pacientes com Transtorno Afetivo Bipolar, apresentando em tempo real flutuações de humor e comportamentos nos participantes através da interação com seus dispositivos móveis. A fenotipagem digital coletada é apresentada aos especialistas em tempo real poderá ajudar a prever alterações no comportamento das pessoas antes que atinjam consequências funcionais extremas. Neste estudo, apresentamos avaliações de usabilidade piloto envolvendo seis usuários não portadores da doença e cinco especialistas para avaliar a percepção que tem do aplicativo. Apresentamos também um estudo longitudinal realizado durante um mês e avaliação em tempo real com especialistas das áreas de psicologia e psiquiatria. Finalmente, apresentamos o potencial de BraPolar no monitoramento remoto de pessoas com transtorno bipolar. / [en] This work addresses the real-time remote monitoring of people with Bipolar Affective Disorder and their interaction with their mobile devices. Some approaches in computer science present experiments for collecting subjective information that can be used by experts for the benefit of people with specific needs. On the other hand, digital phenotyping is a multidisciplinary field of science, used to describe a new approach to monitor user interaction withtheir smartphone and to map disturbances using cell phone sensors. Although, some research assesses the advantages of mobile applications for the treatment of mental illness but no solution has been found in the literature that makes use of digital phenotype analysis and real-time information visualization as state markers and non-invasive therapeutic intervention traits by early detection of changes in patients behavioral patterns. In this dissertation we present BraPolar, an m-Health for remote monitoring of patients with Bipolar Affective Disorder, featuring real-time fluctuations in mood and behavior in participants through interaction with their mobile devices. Collected digital phenotyping presented to real-time specialists can help predict changes in people s behavior before they reach extreme functional consequences. In this study, we present pilot usability assessments involving six non-disease users and five experts to assess their perception of the application. We also present a longitudinal study conducted over a month and real-time evaluation with experts in the fields of psychology and psychiatry. Finally, we present the potential of BraPolar in remote monitoring of people with bipolar disorder.