Deep learning enables sleep staging from photoplethysmogram for patients with suspected sleep apnea

Abstract

Study Objectives Accurate identification of sleep stages is essential in the diagnosis of sleep disorders (e.g. obstructive sleep apnea [OSA]) but relies on labor-intensive electroencephalogram (EEG)-based manual scoring. Furthermore, long-term assessment of sleep relies on actigraphy differentiating only between wake and sleep periods without identifying specific sleep stages and having low reliability in identifying wake periods after sleep onset. To address these issues, we aimed to develop an automatic method for identifying the sleep stages from the photoplethysmogram (PPG) signal obtained with a simple finger pulse oximeter. Methods PPG signals from the diagnostic polysomnographies of susptected OSA patients (n = 894) were utilized to develop a combined convolutional and recurrent neural network. The deep learning model was trained individually for three-stage (wake/NREM/REM), four-stage (wake/N1+N2/N3/REM), and five-stage (wake/N1/N2/N3/REM) classification of sleep. Results The three-stage model achieved an epoch-by-epoch accuracy of 80.1% with Cohen’s κ of 0.65. The four- and five-stage models achieved 68.5% (κ = 0.54), and 64.1% (κ = 0.51) accuracies, respectively. With the five-stage model, the total sleep time was underestimated with a mean bias error (SD) of of 7.5 (55.2) minutes. Conclusion The PPG-based deep learning model enabled accurate estimation of sleep time and differentiation between sleep stages with a moderate agreement to manual EEG-based scoring. As PPG is already included in ambulatory polygraphic recordings, applying the PPG-based sleep staging could improve their diagnostic value by enabling simple, low-cost, and reliable monitoring of sleep and help assess otherwise overlooked conditions such as REM-related OSA.

A Systematic Review of Sensing Technologies for Wearable Sleep Staging

“…The AASM defines five different stages of sleep (wake, N1, N2, N3, and REM), whereas the previous R&K guidelines defined seven stages (wake, S1, S2, S3, S4, and REM). Of all the reviewed articles, only 21 performed classification of all the sleep stages defined by either of these guidelines [ 24 , 39 , 42 , 45 , 50 , 58 , 67 , 72 , 76 , 80 , 88 , 89 , 90 , 93 , 101 , 102 , 105 , 108 , 109 , 117 , 121 ]. Amongst them, all but three [ 24 , 58 , 150 ] used EEG signals for classification, where the difference between sleep stages is known to be most obvious.…”

Detailed Assessment of Sleep Architecture With Deep Learning and Shorter Epoch-to-Epoch Duration Reveals Sleep Fragmentation of Patients With Obstructive Sleep Apnea

“…In a clinical dataset of patients with suspected OSA, the developed method reached at least similar inter-rater reliability (83.8% accuracy, κ = 0.78) [16] as between two manual scorers [20][21][22]. The deep learning-based sleep staging also succeeded in accurately identifying sleep stages from a single EEG channel [16] or even from a photoplethysmography signal [17]. However, the main advantage of the automatic sleep staging over manual scoring is the ability to always score the sleep stages consistently.…”

“…Therefore, we hypothesize that the current sleep staging with non-overlapping 30-second epochs heavily underestimates the extent of sleep fragmentation in patients suffering from OSA. As deep learning-based methods have demonstrated remarkable accuracy in automatic sleep staging [12][13][14][15][16][17], we hypothesize that deep learning offers a unique possibility for providing a more feasible and accurate representation of sleep architecture beyond the non-overlapping 30-second epochs.…”

Proof of principle study: diagnostic accuracy of a novel algorithm for the estimation of sleep stages and disease severity in patients with sleep-disordered breathing based on actigraphy and respiratory inductance plethysmography

“…Korkalainen et al [12] presented a deep learning algorithm for sleep staging based on PPG data in 894 patients with suspected sleep-disordered breathing. They analyzed a three-stage (wake/NREM/REM), a four-stage (wake/N1+N2/N3/REM), and a five-stage (wake/N1/N2/N3/REM) model with strong epoch-by-epoch accuracies of 80.1%, 68.5%, and 64.1%, respectively, resulting in a moderate agreement compared to manual EEG scoring.…”