Von Ralph Dane Marquez Herbuela

Continuous emotion estimation from neurophysiological signals remains challenging due to multimodal heterogeneity, inter-individual variability, and the diffuse nature of emotional experience. Existing methods mainly emphasize signal-level fusion or label-level uncertainty modeling, while providing limited treatment of latent internal-state mechanisms underlying emotional feeling. We propose an interoception-inspired framework for continuous emotion estimation. The framework models emotional feeling through the Brain-inferred Body State, the Interoceptive Error, and their gain-modulated integration, and further introduces Emotional Diffuseness over continuous valence–arousal space to guide non-uniform supervision. Emotional Diffuseness is quantified from contextualized feeling expressions via a Bayesian embedding-to-affective projection, yielding a continuous diffuseness field over affective space. Experiments on the DEAP dataset under leave-one-subject-out evaluation show that the proposed method achieves effective continuous regression and competitive classification performance. Ablation results further confirm the contributions of both explicit interoceptive modeling and diffuseness-guided supervision. These suggest that organizing emotion estimation around internal-state mechanisms and affective diffuseness provides a promising direction for physiological emotion modeling.

Communication interventions have broadened from dialogical meaning-making, assessment approaches, to remote-controlled interactive objects. Yet, interpretation of the mostly pre-or protosymbolic, distinctive, and idiosyncratic movements of children with intellectual disabilities (IDs) or profound intellectual and multiple disabilities (PIMD) using computer-based assistive technology (AT), machine learning (ML), and environment data (ED: location, weather indices and time) remain insufficiently unexplored. We introduce a novel behavior inference computer-based communication-aid AT system structured on machine learning (ML) framework to interpret the movements of children with PIMD/IDs using ED. To establish a stable system, our study aimed to train, cross-validate (10-fold), test and compare the classification accuracy performance of ML classifiers (eXtreme gradient boosting [XGB], support vector machine [SVM], random forest [RF], and neural network [NN]) on classifying the 676 movements to 2, 3, or 7 behavior outcome classes using our proposed dataset recalibration (adding ED to movement datasets) with or without Boruta feature selection (53 child characteristics and movements, and ED-related features). Natural-child-caregiver-dyadic interactions observed in 105 single-dyad video-recorded (30-hour) sessions targeted caregiver-interpreted facial, body, and limb movements of 20 8-to 16-year-old children with PIMD/IDs and simultaneously app-and-sensor-collected ED. Classification accuracy variances and the influences of and the interaction among recalibrated dataset, feature selection, classifiers, and classes on the pooled classification accuracy rates were evaluated using three-way ANOVA. Results revealed that Boruta and NN-trained dataset in class 2 and the non-Boruta SVM-trained dataset in class 3 had >76% accuracy rates. Statistically significant effects indicating high classification rates (>60%) were found among movement datasets: with ED, non-Boruta, class 3, SVM, RF, and NN. Similar trends (>69%) were found in class 2, NN, Boruta-trained movement dataset with ED, and SVM and RF, and non-Boruta-trained movement dataset with ED in class 3. These results support our hypotheses that adding environment data to movement datasets, selecting important features using Boruta, using NN, SVM and RF classifiers, and classifying movements to 2 and 3 behavior outcomes can provide >73.3% accuracy rates, a promising performance for a stable ML-based behavior inference communication-aid AT system for children with PIMD/IDs.

Background: Children with profound intellectual and multiple disabilities (PIMD) or severe motor and intellectual disabilities (SMID) only communicate through movements, vocalizations, body postures, muscle tensions, or facial expressions on a pre- or protosymbolic level. Yet, to the best of our knowledge, there are few systems developed to specifically aid in categorizing and interpreting behaviors of children with PIMD or SMID to facilitate independent communication and mobility. Further, environmental data such as weather variables were found to have associations with human affects and behaviors among typically developing children; however, studies involving children with neurological functioning impairments that affect communication or those who have physical and/or motor disabilities are unexpectedly scarce. Objective: This paper describes the design and development of the ChildSIDE app, which collects and transmits data associated with children’s behaviors, and linked location and environment information collected from data sources (GPS, iBeacon device, ALPS Sensor, and OpenWeatherMap application programming interface [API]) to the database. The aims of this study were to measure and compare the server/API performance of the app in detecting and transmitting environment data from the data sources to the database, and to categorize the movements associated with each behavior data as the basis for future development and analyses. Methods: This study utilized a cross-sectional observational design by performing multiple single-subject face-to-face and video-recorded sessions among purposively sampled child-caregiver dyads (children diagnosed with PIMD/SMID, or severe or profound intellectual disability and their primary caregivers) from September 2019 to February 2020. To measure the server/API performance of the app in detecting and transmitting data from data sources to the database, frequency distribution and percentages of 31 location and environment data parameters were computed and compared. To categorize which body parts or movements were involved in each behavior, the interrater agreement κ statistic was used. Results: The study comprised 150 sessions involving 20 child-caregiver dyads. The app collected 371 individual behavior data, 327 of which had associated location and environment data from data collection sources. The analyses revealed that ChildSIDE had a server/API performance >93% in detecting and transmitting outdoor location (GPS) and environment data (ALPS sensors, OpenWeatherMap API), whereas the performance with iBeacon data was lower (82.3%). Behaviors were manifested mainly through hand (22.8%) and body movements (27.7%), and vocalizations (21.6%). Conclusions: The ChildSIDE app is an effective tool in collecting the behavior data of children with PIMD/SMID. The app showed high server/API performance in detecting outdoor location and environment data from sensors and an online API to the database with a performance rate above 93%. The results of the analysis and categorization of behaviors suggest a need for a system that uses motion capture and trajectory analyses for developing machine- or deep-learning algorithms to predict the needs of children with PIMD/SMID in the future.

[This corrects the article DOI: 10.2196/19034.].

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).

Background: While early detection and effective control of epidemics depend on appropriate surveillance methods, the Philippines bases its dengue fever surveillance system on a passive surveillance method (notifications from barangay/village health centers, municipal or city health offices, hospitals, and clinics). There is no available mHealth (mobile health) app for dengue fever that includes all the appropriate surveillance methods in early detection of disease outbreaks in the country. Objective: This study aimed to evaluate the usability of the Mozzify app in terms of objective quality (engagement, functionality, aesthetics, information) and app subjective and app-specific qualities and compare total app mean score ratings by sociodemographic profile and self and family dengue fever history to see what factors are associated with high app mean score rating among school-based young adult samples and health care professionals. Individual interviews and focus group discussions were also conducted among participants to develop themes from their comments and suggestions to help structure further improvement and future development of the app. Methods: User experience sessions were conducted among participants, and the Mobile Application Rating Scale (MARS) professional and user versions (uMARS) were administered followed by individual interviews and focus group discussions. Descriptive statistical analysis of the MARS and uMARS score ratings was performed. The total app mean score ratings by sociodemographic and dengue fever history using nonparametric mean difference analyses were also conducted. Thematic synthesis was used to develop themes from the comments and suggestions raised in individual interviews and focus group discussions. Results: Mozzify obtained an overall >4 (out of 5) mean score ratings in the MARS and uMARS app objective quality (4.45), subjective (4.17), and specific (4.55) scales among 948 participants (79 health care professionals and 869 school-based samples). Mean difference analyses revealed that total app mean score ratings were not significantly different across ages and gender among health care professionals and across age, income categories, and self and family dengue fever history but not gender (P<.001) among the school-based samples. Thematic syntheses revealed 7 major themes: multilanguage options and including other diseases; Android version availability; improvements on the app’s content, design, and engagement; inclusion of users from low-income and rural areas; Wi-Fi connection and app size concerns; data credibility and issues regarding user security and privacy. Conclusions: With its acceptable performance as perceived by health care professionals and school-based young adults, Mozzify has the potential to be used as a strategic health intervention system for early detection of disease outbreaks in the Philippines. It can be used by health care professionals of any age and gender and by school-based samples of any age, socioeconomic status, and dengue fever history. The study also highlights the feasibility of school-based young adults to use health-related apps for disease prevention.

BACKGROUND Children with profound intellectual and multiple disabilities (PIMD) or severe motor and intellectual disabilities (SMID) only communicate through movements, vocalizations, body postures, muscle tensions, or facial expressions on a pre- or protosymbolic level. Yet, to the best of our knowledge, there are few systems developed to specifically aid in categorizing and interpreting behaviors of children with PIMD or SMID to facilitate independent communication and mobility. Further, environmental data such as weather variables were found to have associations with human affects and behaviors among typically developing children; however, studies involving children with neurological functioning impairments that affect communication or those who have physical and/or motor disabilities are unexpectedly scarce. OBJECTIVE This paper describes the design and development of the ChildSIDE app, which collects and transmits data associated with children’s behaviors, and linked location and environment information collected from data sources (GPS, iBeacon device, ALPS Sensor, and OpenWeatherMap application programming interface [API]) to the database. The aims of this study were to measure and compare the server/API performance of the app in detecting and transmitting environment data from the data sources to the database, and to categorize the movements associated with each behavior data as the basis for future development and analyses. METHODS This study utilized a cross-sectional observational design by performing multiple single-subject face-to-face and video-recorded sessions among purposively sampled child-caregiver dyads (children diagnosed with PIMD/SMID, or severe or profound intellectual disability and their primary caregivers) from September 2019 to February 2020. To measure the server/API performance of the app in detecting and transmitting data from data sources to the database, frequency distribution and percentages of 31 location and environment data parameters were computed and compared. To categorize which body parts or movements were involved in each behavior, the interrater agreement κ statistic was used. RESULTS The study comprised 150 sessions involving 20 child-caregiver dyads. The app collected 371 individual behavior data, 327 of which had associated location and environment data from data collection sources. The analyses revealed that ChildSIDE had a server/API performance >93% in detecting and transmitting outdoor location (GPS) and environment data (ALPS sensors, OpenWeatherMap API), whereas the performance with iBeacon data was lower (82.3%). Behaviors were manifested mainly through hand (22.8%) and body movements (27.7%), and vocalizations (21.6%). CONCLUSIONS The ChildSIDE app is an effective tool in collecting the behavior data of children with PIMD/SMID. The app showed high server/API performance in detecting outdoor location and environment data from sensors and an online API to the database with a performance rate above 93%. The results of the analysis and categorization of behaviors suggest a need for a system that uses motion capture and trajectory analyses for developing machine- or deep-learning algorithms to predict the needs of children with PIMD/SMID in the future.

BACKGROUND While early detection and effective control of epidemics depend on appropriate surveillance methods, the Philippines bases its dengue fever surveillance system on a passive surveillance method (notifications from barangay/village health centers, municipal or city health offices, hospitals, and clinics). There is no available mHealth (mobile health) app for dengue fever that includes all the appropriate surveillance methods in early detection of disease outbreaks in the country. OBJECTIVE This study aimed to evaluate the usability of the Mozzify app in terms of objective quality (engagement, functionality, aesthetics, information) and app subjective and app-specific qualities and compare total app mean score ratings by sociodemographic profile and self and family dengue fever history to see what factors are associated with high app mean score rating among school-based young adult samples and health care professionals. Individual interviews and focus group discussions were also conducted among participants to develop themes from their comments and suggestions to help structure further improvement and future development of the app. METHODS User experience sessions were conducted among participants, and the Mobile Application Rating Scale (MARS) professional and user versions (uMARS) were administered followed by individual interviews and focus group discussions. Descriptive statistical analysis of the MARS and uMARS score ratings was performed. The total app mean score ratings by sociodemographic and dengue fever history using nonparametric mean difference analyses were also conducted. Thematic synthesis was used to develop themes from the comments and suggestions raised in individual interviews and focus group discussions. RESULTS Mozzify obtained an overall >4 (out of 5) mean score ratings in the MARS and uMARS app objective quality (4.45), subjective (4.17), and specific (4.55) scales among 948 participants (79 health care professionals and 869 school-based samples). Mean difference analyses revealed that total app mean score ratings were not significantly different across ages and gender among health care professionals and across age, income categories, and self and family dengue fever history but not gender (P<.001) among the school-based samples. Thematic syntheses revealed 7 major themes: multilanguage options and including other diseases; Android version availability; improvements on the app’s content, design, and engagement; inclusion of users from low-income and rural areas; Wi-Fi connection and app size concerns; data credibility and issues regarding user security and privacy. CONCLUSIONS With its acceptable performance as perceived by health care professionals and school-based young adults, Mozzify has the potential to be used as a strategic health intervention system for early detection of disease outbreaks in the Philippines. It can be used by health care professionals of any age and gender and by school-based samples of any age, socioeconomic status, and dengue fever history. The study also highlights the feasibility of school-based young adults to use health-related apps for disease prevention.