Driver drowsiness monitoring is one of the most demanded technologies for active prevention of severe road accidents. Electroencephalogram (EEG) and several peripheral signals have been suggested for the drowsiness monitoring. However, each type of signal has partial limitations in terms of either convenience or accuracy. Recent emerged concept of in-ear EEG raises expectations due to reduced obtrusiveness. It is yet unclear whether the in-ear EEG is effective enough for drowsiness detection in comparison with on-scalp EEG or peripheral signals. In this work, we evaluated performance of the in-ear EEG in drivers' alertness-drowsiness classification for the first time. Simultaneously, we also tested three peripheral signals including electrocardiogram (ECG), photoplethysmogram (PPG), and galvanic skin response (GSR) which have advantage in convenience of measurement. The classification analysis using the in-ear EEG resulted in high classification accuracy comparable to that of the individual on-scalp EEG channels. The ECG, PPG and GSR showed competitive performance but only when used together in pairwise combinations. Our results suggest that the in-ear EEG would be viable alternative to the single channel EEG or the individual peripheral signals for the drowsiness monitoring.

We investigated the effects of a quality screening method on unconstrained measured signals, including electrocardiogram (ECG), photoplethysmogram (PPG), and ballistocardiogram (BCG) signals, in our collective chair system for smart healthcare. Such an investigation is necessary because unattached or unbound sensors have weaker connections to body parts than do conventional methods. Using the biosignal chair, the physiological signals collected during sessions included a virtual driving task, a physically powered wheelchair drive, and three types of body motions. The signal quality index was defined by the similarity between the observed signals and noise-free signals from the perspective of the cross-correlations of coefficients with appropriate individual templates. The goal of the index was to qualify signals without a reference signal to assess the practical use of the chair in daily life. As expected, motion artifacts have adverse effects on the stability of physiological signals. However, we were able to observe a supplementary relationship between sensors depending on each movement trait. Except for extreme movements, the signal quality and estimated heart rate (HR) remained within the range of criteria usable for status monitoring. By investigating the signal reliability, we were able to confirm the suitability of using the unconstrained biosignal chair to collect real-life measurements to improve safety and healthcare.

Photoplethysmography (PPG) of the foot sole could provide additional health-related information compared with traditional PPG of the finger or wrist. Previously, foot PPG required the procedural binding of a light-emitting diode (LED)-photodetector (PD) pair. We achieved PPG of the foot sole without binding any sensors to the foot while the participant stood in a natural standing position on the testing device. Foot PPG was performed using multiple LED-PD pairs to overcome motion artefacts caused by stabilization. We identified regions of the sole suitable for reliable sensor positioning with optimal LED-PD pairs on the basis of the estimated heart rate (HR) and signal quality index derived by dynamic time warping (wSQI). The first experiment included four participants with direct skin-to-sensor contact, and the results showed a mean HR estimation error of 0.01 beats/min and a wSQI of 0.909. The extended experiment with 53 participants, which involved including a gap between the skin and sensors to consider real-life applications, yielded a mean HR estimation error of 0.638 beats/min and a wSQI of 0.751. Based on the selection ratio of optimal LED-PD pairs, the best region of the sole for PPG was the midfoot, except the medial longitudinal arch. In conclusion, we confirmed that foot PPG using multiple LED-PD pairs is appropriate for HR evaluation and further applications.

Drowsiness while driving can lead to accidents that are related to the loss of perception during emergencies that harm the health. Among physiological signals, brain waves have been used as informative signals for the analyses of behavioral observations, steering information, and other biosignals during drowsiness. We inspected the machine learning methods for drowsiness detection based on brain signals with varying quantities of information. The results demonstrated that machine learning could be utilized to compensate for a lack of information and to account for individual differences. Cerebral area selection approaches to decide optimal measurement locations could be utilized to minimize the discomfort of participants. Although other statistics could provide additional information in further study, the optimized machine learning method could prevent the dangers of drowsiness while driving by considering a transitional state with nonlinear features. Because brain signals can be altered not only by mental fatigue but also by health status, the optimization analysis of the system hardware and software will be able to increase the power-efficiency and accessibility in acquiring brain waves for health enhancements in daily life.

Understanding the factors contributing to depression in farmers is crucial for ensuring their well-being and productivity. To address this issue, our study delves into depression factors among farmers, employing advanced tree-based machine learning (ML) algorithms, specifically focusing on the Category Boosting (CatB) algorithm. Applying the Patient Health Questionnaire-9 (PHQ-9) criteria, 2,446 individuals among 14,810 repondents were classified into depression including mild symptoms. In the classification, CatB achieved an impressive 79.7% accuracy and 81.4% F1 score compared to the other tree-based ensemble models (Random Forest - RF, Extra Trees - ET, and XGBoost - XGB). RF showed the highest sensitivity at 90.0% and the 81.3% F1 score followed by CatB. For the feature importances, the Gini impurity was predominantly used to assess in the RF and ET models. Through the analysis of feature importances, ‘Health’, ‘Sleep time’, ‘Busyness’, ‘Income’, and ‘Frequency of wearing protective gear’ were identified as significant features. These findings underscore the importance of addressing farmers’ mental health support and developing treatment strategies for high-risk individuals in the agricultural sector. Empowering healthcare providers with these tools can lead to more effective interventions, potentially reducing the burden of depression among farmers and enhancing their overall productivity.