0447 ResTNet: A Robust End-to-End Deep Learning Approach to Sleep Staging of Self Applied Somnography Studies

Abstract

Introduction Sleep stage classifications are of central importance when diagnosing various sleep-related diseases. Performing a full PSG recording can be time-consuming and expensive, and often requires an overnight stay at a sleep clinic. Furthermore, the manual sleep staging process is tedious and subject to scorer variability. Here we present an end-to-end deep learning approach to robustly classify sleep stages from Self Applied Somnography (SAS) studies with frontal EEG and EOG signals. This setup allows patients to self-administer EEG and EOG leads in a home sleep study, which reduces cost and is more convenient for the patients. However, self-administration of the leads increases the risk of loose electrodes, which the algorithm must be robust to. The model structure was inspired by ResNet (He, Zhang, Ren, Sun, 2015), which has been highly successful in image recognition tasks. The ResTNet is comprised of the characteristic Residual blocks with an added Temporal component. Methods The ResTNet classifies sleep stages from the raw signals using convolutional neural network (CNN) layers, which avoids manual feature extraction, residual blocks, and a gated recurrent unit (GRU). This significantly reduces sleep stage prediction time and allows the model to learn more complex relations as the size of the training data increases. The model was developed and validated on over 400 manually scored sleep studies using the novel SAS setup. In developing the model, we used data augmentation techniques to simulate loose electrodes and distorted signals to increase model robustness with regards to missing signals and low quality data. Results The study shows that applying the robust ResTNet model to SAS studies gives accuracy > 0.80 and F1-score > 0.80. It outperforms our previous model which used hand-crafted features and achieves similar performance to a human scorer. Conclusion The ResTNet is fast, gives accurate predictions, and is robust to loose electrodes. The end-to-end model furthermore promises better performance with more data. Combined with the simplicity of the SAS setup, it is an attractive option for large-scale sleep studies. Support This work was supported by the Icelandic Centre for Research RANNÍS (175256-0611).

Self‐applied somnography: technical feasibility of electroencephalography and electro‐oculography signal characteristics in sleep staging of suspected sleep‐disordered adults

“…Apart from the detected differences, the overall technical feasibility and morphology of the SAS EEG and EOG signals were on a good level and could be suitable for sleep staging of home PSGs. The findings from previous studies support that, too (Jónsson et al, 2020; Kainulainen et al, 2021; Punjabi et al, 2022).…”

“…As a result, self‐applied EEG electrode sets and devices intended for home sleep monitoring have been introduced (Arnal et al, 2020; Carneiro et al, 2020; Kainulainen et al, 2021; Miettinen, Myllymaa, Muraja‐Murro, et al, 2018; Rusanen et al, 2021). These self‐applied EEG setups have already shown promising results in increasing the accuracy of sleep staging in HSAT to a nearly comparable level with standard PSG (Jónsson et al, 2020; Kalevo et al, 2022; Leino et al, 2022; Levendowski et al, 2017; Rusanen et al, 2023). However, self‐applied EEG is not used clinically and needs more rigorous validation.…”

“…In this study, we focused on the technical evaluation of a self‐applied somnography (SAS) (Figure 1) developed by Nox Medical (Reykjavik, Iceland). Previous studies have considered the success rate and data quality of this setup in a large cohort study (Punjabi et al, 2022), as well as its potential for automatic sleep staging (Jónsson et al, 2020). Moreover, the frontal EEG signal characteristics of this frontal EEG setup have been compared with those of standard PSG in a paediatric cohort (Kainulainen et al, 2021).…”

Robust learning from corrupted EEG with dynamic spatial filtering

“…Code used for data analysis can be found at https://github.com/hubertjb/dynamic-spatial-filtering 4. A recent study reported training a neural network on artificially-corrupted sleep EEG data, with a goal similar to ours[71]; however, this study only appears as a Supplement with little information on the methods and results.…”