Fig. 1: Characteristics of the 43 included validation studies.
From: Evaluating the performance of wearable EEG sleep monitoring devices: a meta-analysis approach
a Electrode positions: Most studies used forehead EEG (n = 26), followed by ear-based devices (n = 16); one study used a neck-worn sensor. b Device type: Prototype devices (n = 24) were more common than commercially available systems (n = 19). c Electrode type: Dry electrodes were used in 24 studies, while 19 employed wet electrodes. d Participants’ health status: Most studies involved healthy participants (n = 29), with fewer including clinical populations (n = 8) or mixed cohorts (n = 6). e Study environment: The majority of studies were conducted in controlled environments (n = 32), while 11 were home-based. f Device scoring method: Machine learning (n = 11), manual scoring (n = 10), and proprietary algorithms (n = 9) were the most common, with some studies using deep learning (n = 8) or multiple methods (n = 5). (g) Age distribution: Most studies involved adults in their mid-20s to mid-30s. Mean age, standard deviation (black lines), and range (white bars) are shown. Older adults and adolescents were underrepresented.
