Mamba-based deep learning approach for sleep staging on a wireless multimodal wearable system without electroencephalography

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: 3-class balanced accuracy: 84.02%, 4-class: 75.30%, 5-class: 65.11%

• Primary Outcome: Sleep stage classification accuracy using wearable sensors without EEG

• Nnt Or Sensitivity: Cohen's κ: 72.89% (3-class), 61.51% (4-class), 53.23% (5-class)

• Confidence Interval: Not reported

Clinical Application

Requires specialised wearable device (ANNE One) and technical expertise for model implementation Could complement PBS-funded sleep studies, potentially reducing wait times for sleep clinic assessments. Would require TGA approval for clinical use. May align with RACGP guidelines for sleep disorder screening in primary care. Adults requiring sleep monitoring, particularly those unable to tolerate traditional polysomnography

Abstract

STUDY OBJECTIVES: We investigate a Mamba-based deep learning approach for sleep staging on signals from ANNE One (Sibel Health, Chicago, IL), a non-intrusive dual-module wireless wearable system measuring chest electrocardiography, triaxial accelerometry, chest temperature, and finger photoplethysmography and finger temperature. METHODS: We obtained wearable sensor recordings from 357 adults undergoing concurrent polysomnography at a tertiary care sleep lab. Each polysomnography recording was manually scored, and these annotations served as ground truth labels for training and evaluation of our models. Polysomnography and wearable sensor data were automatically aligned using their electrocardiography channels with manual confirmation by visual inspection. We trained a Mamba-based recurrent neural network architecture on these recordings. Ensembling of model variants with similar architectures was performed. RESULTS: After ensembling, the model attains a 3-class (wake, non-rapid eye movement sleep, rapid eye movement sleep) balanced accuracy of 84.02 per cent, F1 score of 84.23 per cent, Cohen's κ of 72.89 per cent, and a Matthews correlation coefficient (MCC) score of 73.00 per cent; a 4-class (wake, light NREM [N1/N2], deep NREM [N3], REM) balanced accuracy of 75.30 per cent, F1 score of 74.10 per cent, Cohen's κ of 61.51 per cent, and MCC score of 61.95 per cent; a 5-class (wake, N1, N2, N3, REM) balanced accuracy of 65.11 per cent, F1 score of 66.15 per cent, Cohen's κ of 53.23 per cent, MCC score of 54.38 per cent. CONCLUSIONS: Our Mamba-based deep learning model can successfully infer major sleep stages from the ANNE One, a wearable system without electroencephalography, and can be applied to data from adults attending a tertiary care sleep clinic.

References

1. 1.Zhang, A. H., He-Mo, A., Yin, R. F., Li, C., Tang, Y., Gurve, D., van der Horst, V., Buchman, A. S., Ghahjaverestan, N. M., Goubran, M., Wang, B., & Lim, A. S. P. (2026). Mamba-based deep learning approach for sleep staging on a wireless multimodal wearable system without electroencephalography. Sleep. https://doi.org/10.1093/sleep/zsag022