Personalized mental stress detection with self-organizing map: From laboratory to the field

Personalized Stress Detection Using Biosignals from Wearables: A Scoping Review

Stress is a natural yet potentially harmful aspect of human life, necessitating effective management, particularly during overwhelming experiences. This paper presents a scoping review of personalized stress detection models using wearable technology. Employing the PRISMA-ScR framework for rigorous methodological structuring, we systematically analyzed literature from key databases including Scopus, IEEE Xplore, and PubMed. Our focus was on biosignals, AI methodologies, datasets, wearable devices, and real-world implementation challenges. The review presents an overview of stress and its biological mechanisms, details the methodology for the literature search, and synthesizes the findings. It shows that biosignals, especially EDA and PPG, are frequently utilized for stress detection and demonstrate potential reliability in multimodal settings. Evidence for a trend towards deep learning models was found, although the limited comparison with traditional methods calls for further research. Concerns arise regarding the representativeness of datasets and practical challenges in deploying wearable technologies, which include issues related to data quality and privacy. Future research should aim to develop comprehensive datasets and explore AI techniques that are not only accurate but also computationally efficient and user-centric, thereby closing the gap between theoretical models and practical applications to improve the effectiveness of stress detection systems in real scenarios.

A Deep Learning-Based Platform for Workers' Stress Detection Using Minimally Intrusive Multisensory Devices

The advent of Industry 4.0 necessitates substantial interaction between humans and machines, presenting new challenges when it comes to evaluating the stress levels of workers who operate in increasingly intricate work environments. Undoubtedly, work-related stress exerts a significant influence on individuals' overall stress levels, leading to enduring health issues and adverse impacts on their quality of life. Although psychological questionnaires have traditionally been employed to assess stress, they lack the capability to monitor stress levels in real-time or on an ongoing basis, thus making it arduous to identify the causes and demanding aspects of work. To surmount this limitation, an effective solution lies in the analysis of physiological signals that can be continuously measured through wearable or ambient sensors. Previous studies in this field have mainly focused on stress assessment through intrusive wearable systems susceptible to noise and artifacts that degrade performance. One of our recently published papers presented a wearable and ambient hardware-software platform that is minimally intrusive, able to detect human stress without hindering normal work activities, and slightly susceptible to artifacts due to movements. A limitation of this system is its not very high performance in terms of the accuracy of detecting multiple stress levels; therefore, in this work, the focus was on improving the software performance of the platform, using a deep learning approach. To this purpose, three neural networks were implemented, and the best performance was achieved by the 1D-convolutional neural network with an accuracy of 95.38% for the identification of two levels of stress, which is a significant improvement over those obtained previously.

Investigation of different ML approaches in classification of emotions induced by acute stress

Background

Machine learning is becoming a common tool in monitoring emotion. However, methodological studies of the processing pipeline are scarce, especially ones using subjective appraisals as ground truth.

New method

A novel protocol was used to induce cognitive load and physical discomfort, and emotional dimensions (arousal, valence, and dominance) were reported after each task. The performance of five common ML models with a versatile set of features (physiological features, task performance data, and personality trait) was compared in binary classification of subjectively assessed emotions.

Results

The psychophysiological responses proved the protocol was successful in changing the mental state from baseline, also the cognitive and physical tasks were different. The optimization and performance of ML models used for emotion detection were evaluated. Additionally, methods to account for imbalanced classes were applied and shown to improve the classification performance.

Comparison with existing methods

Classification of human emotional states often assumes the states are determined by the stimuli. However, individual appraisals vary. None of the past studies have classified subjective emotional dimensions with a set of features including biosignals, personality and behavior.

Conclusion

Our data represent a typical setup in affective computing utilizing psychophysiological monitoring: N is low compared to number of features, inter-individual variability is high, and class imbalance cannot be avoided. Our observations are a) if possible, include features representing physiology, behavior and personality, b) use simple models and limited number of features to improve interpretability, c) address the possible imbalance, d) if the data size allows, use nested cross-validation.

Optimizing future well-being with artificial intelligence: self-organizing maps (SOMs) for the identification of islands of emotional stability

In this article, we present a deep learning model of human psychology that can predict one’s current age and future well-being. We used the model to demonstrate that one’s baseline well-being is not the determining factor of future well-being, as posited by hedonic treadmill theory. Further, we have created a 2D map of human psychotypes and identified the regions that are most vulnerable to depression. This map may be used to provide personalized recommendations for maximizing one’s future well-being.

Machine Learning for personalised stress detection: Inter-individual variability of EEG-ECG markers for acute-stress response

Stress appears as a response for a broad variety of physiological stimuli. It does vary among individuals in amplitude, phase and frequency. Thus, the necessity for personalised diagnosis is key to prevent stress-related diseases. In order to evaluate stress levels, a multi-sensing system is proposed based on non-invasive EEG and ECG signals. A target population of 24 individuals which age range between 18-23 years old are intentionally exposed to control-induced stress tests while EEG and ECG are simultaneously recorded. The acquired signals are processed by using semisupevised Machine Learning techniques as those provide a patient-specific approach due to key characteristics such as adaptiveness and robustness. In here, a stress metric is proposed that jointly with each individual medical history provide mechanisms to prevent and avoid possible chronic-health issues for individuals whom are more sensitive to stressors. Finally, supervised learning techniques are used to classify the obtained featured clusters to evaluate specific and general subject models in order to pave the way for real time stress monitoring.

Ultra-Short Window Length and Feature Importance Analysis for Cognitive Load Detection from Wearable Sensors

Human cognitive capabilities are under constant pressure in the modern information society. Cognitive load detection would be beneficial in several applications of human–computer interaction, including attention management and user interface adaptation. However, current research into accurate and real-time biosignal-based cognitive load detection lacks understanding of the optimal and minimal window length in data segmentation which would allow for more timely, continuous state detection. This study presents a comparative analysis of ultra-short (30 s or less) window lengths in cognitive load detection with a wearable device. Heart rate, heart rate variability, galvanic skin response, and skin temperature features are extracted at six different window lengths and used to train an Extreme Gradient Boosting classifier to detect between cognitive load and rest. A 25 s window showed the highest accury (67.6%), which is similar to earlier studies using the same dataset. Overall, model accuracy tended to decrease as the window length decreased, and lowest performance (60.0%) was observed with a 5 s window. The contribution of different physiological features to the classification performance and the most useful features that react in short windows are also discussed. The analysis provides a promising basis for future real-time applications with wearable sensors.