E4 wristband | Real-time physiological signals | Wearable PPG, EDA, Temperature, Motion sensors

Exploiting physiological data to detect virtual reality sickness

Researchers at IRT b-com, France, explored the use of physiological signals from noninvasive wearable devices like the E4 to detect virtual reality sickness. Their method involved training Machine Learning models using data from the E4 to detect when participants felt sick while playing video games for 30 minutes, in addition to self-reporting. Results from the study showed up to 91% accuracy in detecting VR sickness when using the E4, paving the way for adapting VR video games to players’ sickness levels using real-time feedback.

Classifying emotions through multimodal signal recordings

Gloria Cosoli and her team at Università Politecnica Delle Marche (Italy) researched the possibility of utilizing data from multiple E4 sensors to identify and classify emotions. Data were recorded from 7 healthy participants at rest, with the participants wearing the Empatica E4 on the dominant wrist while listening to 1-minute audio recordings as stimuli. Results showed up to 75% accuracy across all the evaluation metrics, and the data from the algorithms trained in this research have been validated on a publicly available dataset (WESAD).

Measuring stress intensity through Machine Learning methods

In this study, Pekka Sirtola et al. collected data from the E4, analyzing it using different Machine Learning models to determine the feasibility of detecting and identifying different levels of stress. Participants were asked to drive a car in stress and non-stress situations while wearing the E4, and regression and classification ML models were compared during data analysis. Their results show that regression models outperform classification models in stress detection, and unlike the latter, can also be used to identify different levels of stress.

How Dutch and German visitors experience an exhibit of Second World War stories

Researchers used the social identity theory framework to assess differences in emotional reactions of Dutch and German visitors to stories of the Second World War, as presented at a Dutch museum exhibit. E4 wristbands were worn by visitors to measure emotional reactions using physiological signals of heart rate and heart rate variability, in addition to self-reporting via a tablet-administered intake questionnaire. It was found that patterns in the physiological and self-report data differed, and that generally participants did not simply categorize themselves with either national or human identities of characters based on what their respective stories emphasized.

Exploring play as a healing factor in hospitalized children

Using the Empatica E4 and SDK, researchers at Copenhagen University Hospital developed a method that uses play to help children overcome difficult experiences in a hospital environment. The solution consists of a physical teddy bear and an E4 wristband, connected via a custom-designed tablet/smartphone application with a virtual teddy bear that the child can interact with. The children are asked to act as caregivers for the teddy bear while interacting with the app, and help the teddy bear through the treatment. This shift in focus and perspective for the child increased emotional stability, with a calming effect.

E4 Bytes: Predicting aggressive outbursts in people with autism a minute in advance

Using the E4, Northeastern behavioral scientist Matthew Goodwin and his team have created an algorithm that can predict aggressive outbursts in people with autism by monitoring physiological indicators of stress. By analyzing the changes in physiology that occurred around each episode, the algorithm developed by the researchers could predict an aggressive outburst a minute in advance with 84 percent accuracy. Even a minute's warning can be crucial in helping caregivers prevent an aggressive outburst, showing there is true potential in the use of wearables and AI to alert caregivers and help mitigate their emergence, occurrence, or impact.

Electrophysiological evidence of emotional engagement during a roller-coaster ride with VR Add-On

This research evaluated the methodological feasibility and usefulness of ambulatory recordings of skin conductance (SC) responses during a tourism experience. The goal was to measure emotions accurately while experiences unfolded in time. Skin conductance (SC) was recorded with E4 wristbands in participants while they experienced a roller coaster ride with or without a Virtual Reality (VR) headset. Through the collected data, the researchers found that SC response time series were meaningfully related to the different ride elements, establishing psychophysiological measurements as a new avenue for understanding how hospitality, tourism & leisure experiences dynamically develop over time.

Developing a model to predict migraine attacks using biosignals

Researchers at The University of Oulu used the E4 in a study on the early detection of migraine attacks via human-measured biosignals. The aim was to develop a predictive model that assists individuals in taking their medication on time, preventing painful attacks. The E4 was used to collect sleep data and, altogether, 110 features were extracted from each night’s biosignals, and used to train machine learning models. The experiments showed that early symptoms for migraine are highly personal, and that using personal recognition models the accuracy for detecting attacks one night prior is over 84%.

Predicting sepsis in hospitalized patients

The detection of fever has played a central part in patient monitoring. Nursing observations are often taken as part of standard vital signs, the frequency depending on patient acuity. However, this is time-consuming and may miss important spikes in temperature suggesting incipient or unrecognized sepsis. In a study conducted at The Royal Melbourne Hospital, the E4 was used to monitor a series of admitted patients. The researchers found that a temperature of 37.5 or more reliably identified patients with infection, concluding that peripheral temperature recording has the potential to provide an early indication for patients with sepsis.

Detecting moments of stress in real-world settings

Researchers at the University of Salzburg, University Hospital Zurich, Harvard University, University of Groningen, and the University of Birmingham, aimed to introduce a new algorithm that leverages the capabilities of wearable physiological sensors to detect moments of stress (MOS). They proposed a rule-based algorithm based on galvanic skin response and skin temperature, which they measured with the E4. The new algorithm combines empirical findings with expert knowledge to ensure transferability between laboratory settings and real-world field studies. Results show that the algorithm detects MOS with 84% accuracy, showing high correlations between measured (by wearable sensors), reported (by questionnaires and eDiary entries), and recorded (by video) stress events.

Wearables and the quantified self: Systematic benchmarking of physiological sensors

Though wearable sensors are increasingly used in research, it is often not clear how accurate their measurements are compared to those from well-calibrated, high-end laboratory equipment. Researchers at the University of Salzburg and Harvard demonstrated an approach to quantify the accuracy of wearables, including the E4, in comparison to laboratory sensors. The benchmarked wearables provided physiological measurements such as heart rate and inter-beat interval with an accuracy close to that of the professional high-end sensor. The accuracy varies more for other parameters, such as galvanic skin response, yet the E4 demonstrated extraordinary stability and high quality throughout.