Real-Time Stress Level Feedback from Raw Ecg Signals for Personalised, Context-Aware Applications Using Lightweight Convolutional Neural Network Architectures

Affect and stress detection based on feature fusion of LSTM and 1DCNN

The impact of emotions on health, especially stress, is receiving increasing attention. It is important to provide a non-invasive affect detection system that can be continuously monitored for a long period of time. Multi-sensor fusion strategies can better improve the performance of affect detection models, but there are also problems such as insufficient feature extraction and poor spatiotemporal feature fusion. Therefore, this study proposes a feature-level fusion method based on long short-term memory and one-dimensional convolutional neural network to extract temporal and spatial features of electrocardiogram, electromyogram, electrical activity, temperature, accelerator and response data, respectively, and then fuse them in a summation fashion for affect and stress detection. In particular, we added the tanh activation function before feature fusion, which can improve the model's performance. We used the wearable affect and stress detection dataset to train the model, which includes three different emotion states (neutral, stress, and amusement) for three-class emotion classification with accuracy and F1-scores of 87.82% and 86.68%, respectively. Due to the importance of stress, we also studied binary classification for stress detection, where neutral and amusement were combined as non-stress, with accuracy and F1-scores of 94.9% and 94.98%, respectively. The performance of the proposed model outperforms other control models and can effectively improve the performance of affect and stress detection, and promote medical care, health care and elderly care.

ECG-Based Stress Detection and Productivity Factors Monitoring: The Real-Time Production Factory System

Productivity and production quality have become primary goals for the success of companies in all industrial and manufacturing sectors. Performance in terms of productivity is influenced by several factors including machinery efficiency, work environment and safety conditions, production processes organization, and aspects related to workers' behavior (human factors). In particular, work-related stress is among the human factors that are most impactful and difficult to capture. Thus, optimizing productivity and quality in an effective way requires considering all these factors simultaneously. The proposed system aims to detect workers' stress and fatigue in real time using wearable sensors and machine learning techniques and also integrate all data regarding the monitoring of production processes and the work environment into a single platform. This allows comprehensive multidimensional data analysis and correlation research, enabling organizations to improve productivity through appropriate work environments and sustainable processes for workers. The on-field trial demonstrated the technical and operational feasibility of the system, its high degree of usability, and the ability to detect stress from ECG signals exploiting a 1D Convolutional Neural Network (accuracy 88.4%, F1-score 0.90).

A clinical and technical methodological review on stress detection and sleep quality prediction in an academic environment

BACKGROUND

Mental health in recent times is a much talked about topic and its effects on the sleep health of the students are said to result in long-term health issues if not identified and resolved. Students who are subjected to psychological stress have often been reported to have lower sleep quality which together has affected the academic performance of the students.

OBJECTIVE

While stress has its adverse effect on students'quality of sleep, an effort is also made to identify standard techniques and tools to automatically assess stress levels and sleep quality in a non-invasive environment among students only. This article mainly focuses on the Clinical and technical methodology employed in stress level detection and sleep quality prediction among students.

METHODS

This study was conducted by examining all research studies conducted in the past with respect to students in an academic setting from year 2000 to early 2022. The papers under study where finalised based on different methodologies involved in stress level detection and sleep quality prediction considering both in unimodal and multimodal measurements.

RESULTS

While questionnaires and physiological signals are used as a standard measuring tool, it is mostly used in a unimodal environment to measure students' mental stress or sleep quality in academic settings.

CONCLUSION

This paper describes in detail the clinical aspect of the association between mental stress, sleep quality, and academic performance in students followed by technical aspects to analyse the stress levels and sleep quality both qualitatively and quantitatively in an academic environment.

Recent Advances in Multiplexed Wearable Sensor Platforms for Real-Time Monitoring Lifetime Stress: A Review

Researchers are interested in measuring mental stress because it is linked to a variety of diseases. Real-time stress monitoring via wearable sensor systems can aid in the prevention of stress-related diseases by allowing stressors to be controlled immediately. Physical tests, such as heart rate or skin conductance, have recently been used to assess stress; however, these methods are easily influenced by daily life activities. As a result, for more accurate stress monitoring, validations requiring two or more stress-related biomarkers are demanded. In this review, the combinations of various types of sensors (hereafter referred to as multiplexed sensor systems) that can be applied to monitor stress are discussed, referring to physical and chemical biomarkers. Multiplexed sensor systems are classified as multiplexed physical sensors, multiplexed physical-chemical sensors, and multiplexed chemical sensors, with the effect of measuring multiple biomarkers and the ability to measure stress being the most important. The working principles of multiplexed sensor systems are subdivided, with advantages in measuring multiple biomarkers. Furthermore, stress-related chemical biomarkers are still limited to cortisol; however, we believe that by developing multiplexed sensor systems, it will be possible to explore new stress-related chemical biomarkers by confirming their correlations to cortisol. As a result, the potential for further development of multiplexed sensor systems, such as the development of wearable electronics for mental health management, is highlighted in this review.