Strategies for Reliable Stress Recognition: A Machine Learning Approach Using Heart Rate Variability Features.

Stress recognition, particularly using machine learning (ML) with physiological data such as heart rate variability (HRV), holds promise for mental health interventions. However, limited datasets in affective computing and healthcare research can lead to inaccurate conclusions regarding the ML model performance. This study employed supervised learning algorithms to classify stress and relaxation states using HRV measures. To account for limitations associated with small datasets, robust strategies were implemented based on methodological recommendations for ML with a limited dataset, including data segmentation, feature selection, and model evaluation. Our findings highlight that the random forest model achieved the best performance in distinguishing stress from non-stress states. Notably, it showed higher performance in identifying stress from relaxation (F1-score: 86.3%) compared to neutral states (F1-score: 65.8%). Additionally, the model demonstrated generalizability when tested on independent secondary datasets, showcasing its ability to distinguish between stress and relaxation states. While our performance metrics might be lower than some previous studies, this likely reflects our focus on robust methodologies to enhance the generalizability and interpretability of ML models, which are crucial for real-world applications with limited datasets.

Co-Designing Auditory Navigation Solutions for Traveling as a Blind Individual During the COVID-19 Pandemic.

A group co-design was held in March 2021 with six blind and low-vision individuals (BLIs) from the United States. Participants were asked to discuss problems related to travel during the COVID-19 pandemic and make recommendations for possible solutions. Two probes (prototypes) of a non-visual neighborhood travel map and a non-visual COVID-19 choropleth map (a map using colors or sounds over each state to represent different values) of COVID-19 state data were shown to participants for inspiration. The participants expressed that COVID-19 had significantly increased their apprehension and discomfort associated with activities such as venturing outside, traveling, engaging with strangers, communicating, adapting to changes in familiar environments, and wearing masks. Participants gravitated towards the need for information the probes provided, and made a number of observations and recommendations for improvement. They wanted more detailed geo-referenced COVID-19 data (including by county), information related to voting, a mobile app, and more detailed building information, such as doors on the travel map.

Short-Term Effects of Heart Rate Variability Biofeedback on Working Memory.

Drawing upon the well-documented impact of long-term heart rate variability biofeedback (HRVB) on psychophysiological responses, this study seeks to explore the short-term effects arising from a single HRVB session during and after paced breathing exercise. The research aligns with the neurovisceral integration model, emphasizing the link between heart rate variability (HRV) levels and cognitive performance. Therefore, a randomized controlled trial employing a between-subjects design was conducted with 38 participants. Each participant was assigned to either the paced breathing intervention group or the spontaneous breathing control group. The study assessed various parameters such as cardiac vagal tone, evaluated through vagally mediated HRV measures, and working memory, measured using the N-back task. Additionally, participants' affective states were assessed through self-reported questionnaires, specifically targeting attentiveness, fatigue, and serenity. The results notably reveal enhancements in the working memory task and an elevated state of relaxation and attention following the HRVB session, as evidenced by higher averages of correct responses, serenity and attentiveness scores. However, the findings suggest that this observed improvement is not influenced by changes in cardiac vagal tone, as assessed using a simple mediation analysis. In conclusion, this study presents promising insights into the impact of a single HRVB session, laying the foundation for future research advancements in this domain.

EVALUATION OF A NON-VISUAL AUDITORY CHOROPLETH AND TRAVEL MAP VIEWER.

The auditory virtual reality interface of Audiom, a web-based map viewer, was evaluated by thirteen blind participants. In Audiom, the user is an avatar that navigates, using the arrow keys, through geographic data, as if they are playing a first-person, egocentric game. The research questions were: What will make blind users want to use Audiom maps? And Can participants demonstrate basic acquisition of spatial knowledge after viewing an auditory map? A dynamic choropleth map of state-level US COVID-19 data, and a detailed OpenStreetMap powered travel map, were evaluated. All participants agreed they wanted more maps of all kinds, in particular county-level COVID data, and they would use Audiom once some bugs were fixed and their few recommended features were added. Everyone wanted to see Audiom embedded in their existing travel and mapping applications. All participants were able to answer a question evaluating spatial knowledge. Participants also agreed this spatial information was not available in existing applications.