Even though more individuals are becoming aware of how important sleep is, every year sees an increase in the number of people who experience inadequate sleep. Polysomnography (PSG) with a variety of sensors is the gold-standard sleep assessment method for detecting abnormal sleep patterns. However, many people with sleep disorders go untreated as a result of PSG's high price and limited availability. Recent advancements in wearable sensors and electronics make it possible to continuously monitor sleep at home while overcoming PSG's limitations (Dr. Hobson).
During a night of sleep, the brain is crucial in controlling a variety of biochemical and physiological processes and changes that take place both in our body and in the brain. The most accurate indicator of the five stages of sleep (W, N1, N2, N3, and REM) and any irregularities occurring in sleep disorders is frequently changes in brain activity. As a result, research and assessments of sleep start by observing variations in brain activity while electroencephalograms (EEGs) record brain waves. Therefore, any sleep monitoring system that wants to thoroughly investigate sleep must include the measurement of EEG. For instance, standard PSG allocates the most electrodes to the measurement of EEG on the scalp (Dr. Berry).
Making electrodes that can measure EEG from the scalp, which is typically covered by hair, is the biggest design problem for sleep monitoring devices that may be used outside hospitals. Conductive gels are frequently used in traditional PSG to pierce the hair and create an electrical conduit between the scalp and electrodes that are held in place by hair caps. However, it can be difficult for people to use conductive gel and to set up electrodes with a hair cap, and it takes a lot of time and effort to set everything up and clean it up. Due to its physical proximity to the brain and the presence of a smooth, flat skin surface where moist or dry, many modern technologies attempt to use the forehead as an alternate location for EEG measurement (Dr. Arai).
For instance, the Advanced Brain Monitoring (California, USA) company unveiled their Sleep Profiler home sleep monitoring system, which offers wireless sleep monitoring in a headband platform with three Frontopolar EEG electrodes on the forehead. Additionally, a Photoplethysmography (PPG) sensor, microphone, and triaxial accelerometer are included in this gadget to track body movement, snoring, and pulse rate all at once. The integrated software primarily examines the power spectrum of brain waves in conjunction with other data to identify specific events, such as the sleep spindle and K-complex, and to automatically categorize the stages of sleep. When compared to stages assessed by five additional sleep technologists over 33,635 epochs, the automated five-class sleep staging results indicate Cohen's kappa coefficient of 0.67 on average, whereas the average kappa score among the five sleep technologists was 0.70. Additionally, the sleep profiler evaluates the variability of numerous sleep metrics and events from one night to the next, such as sleep time, sleep latency, waking after sleep onset (WASO), etc (Dr. Arnal).
Despite the hair's barrier, various innovative technologies and studies have made strides toward creating electrodes with novel materials and structures to make scalp EEG measurements trustworthy and practical. A commercially available product called the Dreem Headband measures seven EEG derivations from three dry electrodes placed on the forehead and two dry electrodes placed at the back of the head with soft, flexible silicone protrusions that make contact with the scalp by piercing the hairs. Additionally, this wireless headband system has a PPG sensor for heart rate (HR) monitoring and a 3-axis accelerometer for movement, position, and respiration rate monitoring.
With 25 patients over the course of a single night of sleep research, this device was compared with PSG, which demonstrated their high correlation of recorded brain waves, HR, respiration rate, and respiration rate variability. Additionally, their automated sleep staging deep learning method demonstrated a Cohen's kappa coefficient of 0.748 on average with the other five scorers when the average kappa score among the five scorers was 0.798. This algorithm was taught using a total of 423 previously observed datasets. Total sleep time (TST), sleep onset latency (SOL), and WASO are only a few of the parameters relating to sleep quality that are provided by this Headband. More recently, tests have been conducted on several body areas other than the forehead to measure EEG with less disruption and interruption to normal sleep behavior, including in-Ear is one of the promising regions.
It is also possible to use an in-ear EEG measurement device with two fabric-based EEG electrodes built within a memory-foam substrate. According to the author, memory foam's special mechanical properties enable a secure skin-electrode connection, a snug fit for the user's ear, and an effective reduction of signal artifacts from pulsatile ear canal movements brought on by blood vessel pulsing. The comparison study between EEG and commercial PSG measurements over 21 people demonstrates the latter's capacity to identify slow-wave sleep (SWS), measure sleep latency, and perform automated five-stage sleep scoring. In-ear EEG automated scoring yields a Cohen's kappa coefficient of 0.61 as opposed to manual scoring and 0.79 with scalp EEG (Dr. Sterr).
EEGGrid is another novel method with EEG measurement near the ear. A flexible, thin adhesive strip with ten integrated electrodes that inserts behind the ear serves as the platform for an EEG measurement. By demonstrating that the user may finish the measurement setup by themselves in about 20 minutes, the author highlights the convenience of the cEEGrid. With the assistance of one skilled sleep specialist, the comparison PSG setup took 45 minutes to complete. cEEGrid and PSG setup data were manually assessed by trained scorers to validate its performance, and the average Cohen's kappa coefficient between scores from the two systems was 0.42.
The researcher also displays another kind of ear-mounted EEG measurement equipment, this time with two earplugs mounting a total of 12 dry electrodes. To evaluate the system's performance and convenience, 80 sleep measurements were taken simultaneously by 20 individuals using a PSG setup. As 19 of the 20 individuals claimed that the system had little to no negative effects on their sleep, a high level of convenience and comfort was evident. In most cases, they could use the equipment on their ears unattended. Additionally, a machine learning technique was used to train an automated scoring algorithm with the sleep EEG data acquired by the device, which demonstrated Cohen's kappa values of 0.81 in five-stage sleep scoring when compared to the PSG setup measurement done manually.