At-home sleep monitoring using generic ear-EEG

Abstract

Introduction: A device comprising two generic earpieces with embedded dry electrodes for ear-centered electroencephalography (ear-EEG) was developed. The objective was to provide ear-EEG based sleep monitoring to a wide range of the population without tailoring the device to the individual.

Methods: To validate the device ten healthy subjects were recruited for a 12-night sleep study. The study was divided into two parts; part A comprised two nights with both ear-EEG and polysomnography (PSG), and part B comprised 10 nights using only ear-EEG. In addition to the electrophysiological measurements, subjects filled out a questionnaire after each night of sleep.

Results: The subjects reported that the ear-EEG system was easy to use, and that the comfort was better in part B. The performance of the system was validated by comparing automatic sleep scoring based on ear-EEG with PSG-based sleep scoring performed by a professional trained sleep scorer. Cohen's kappa was used to assess the agreement between the manual and automatic sleep scorings, and the study showed an average kappa value of 0.71. The majority of the 20 recordings from part A yielded a kappa value above 0.7. The study was compared to a companioned study conducted with individualized earpieces. To compare the sleep across the two studies and two parts, 7 different sleeps metrics were calculated based on the automatic sleep scorings. The ear-EEG nights were validated through linear mixed model analysis in which the effects of equipment (individualized vs. generic earpieces), part (PSG and ear-EEG vs. only ear-EEG) and subject were investigated. We found that the subject effect was significant for all computed sleep metrics. Furthermore, the equipment did not show any statistical significant effect on any of the sleep metrics.

Discussion: These results corroborate that generic ear-EEG is a promising alternative to the gold standard PSG for sleep stage monitoring. This will allow sleep stage monitoring to be performed in a less obtrusive way and over longer periods of time, thereby enabling diagnosis and treatment of diseases with associated sleep disorders.

Keywords: ear-EEG; electroencephalography; home recording; long-term sleep monitoring; sleep monitoring.

FIGURE 1

(Left) Generic earpieces with mounted electrodes. The placement of the three main features (ear canal, tail, and main body) are specified for each earpiece. The generic earpiece was molded in soft silicone in one piece to ensure good comfort. Cable relief was incorporated in the design to reduce movement artifacts from cable pulling. Also, a formed tube was designed to guide the cables behind the ear to further reduce cable movement, to increase comfort and to make the earpiece more discrete. EL1, EL2, ER1, and Er2: data electrodes, D1 and D2 ground electrodes, ref: reference electrode. (Right) Earpiece mounted in the ear.

FIGURE 2

Overview of the data. sG refers to data from the study with generic earpieces, and sC to data from the study with custom earpieces. Both datasets have a Part A and a Part B. Part A were recorded with both polysomnography (PSG) and ear-centered electroencephalography (ear-EEG), whereas Part B were recorded from ear-EEG only.

FIGURE 3

The participants rated the comfort and ease-of-use of the device after each night. In Part A, they slept wearing both the polysomnography (PSG) setup and the earpieces, whereas in Part B they only wore the earpieces. This figure shows a summary of their ratings. The numbers in the bars reflect the number of responses (20 in total for Part A and 100 in total for Part B). Inter subject variation (Inter Std) and Intra subject variation (Intra Std) are shown for each question-part combination.

FIGURE 4

Percentage of rejected data is illustrated for each noise category for each electrode. The overall rejection percentage was 10.2% before and 4.4% after channel selection. EL1, left electrode 1; EL2, left electrode 2; ER1, right electrode 1; ER2, right electrode 2.

FIGURE 5

Confusion matrix for sleep scoring with the sCGsG cross-validation method.

FIGURE 6

(Left panels) Histograms of kappa values for sG recordings (cross-validation scheme: sGsG: purple, sCsG: blue, sCGsG: green). Dashed lines represent the average of each method. For the sGsG and sCGsG schemes, the majority of the recordings yielded a kappa value above 0.7. (Right panel) Kappa values for each subject for all three cross-validation schemes. Each point represents one recording.

FIGURE 7

(Left) Confidence versus kappa value for the sG.A recordings. The 25th, 50th, and 75th percentiles of the kappa values and the corresponding confidences are shown with dashed lines. (Center) Histogram of the confidence values for the sG.B recordings. (Right) Confidence values for each recording. The 25th, 50th, and 75th percentile confidence values are presented with dashed lines in all plots.

FIGURE 8

The distribution of the value of the sleep metrics (REMfr, N3fr, SE, NREM to NREM, NREM to REM, REM to REM, and REM to NREM) for sC.A, sC.B, sG.A, and sG.B datasets show a large overlap and no clear equipment or part differences.