Sleep-Deep-Net learns sleep wake scoring from the end-user and completes each record in their style

Abstract

Sleep-wake scoring is a time-consuming, tedious but essential component of clinical and pre-clinical sleep research. Sleep scoring is even more laborious and challenging in rodents due to the smaller EEG amplitude differences between states and the rapid state transitions which necessitate scoring in shorter epochs. Although many automated rodent sleep scoring methods exist, they do not perform as well when scoring new data sets, especially those which involve changes in the EEG/EMG profile. Thus, manual scoring by expert scorers remains the gold-standard. Here we take a different approach to this problem by using a neural network to accelerate the scoring of expert scorers. Sleep-Deep-Net (SDN) creates a bespoke deep convolution neural network model for individual electroencephalographic or local-field-potential records via transfer learning of GoogleNet, by learning from a small subset of manual scores of each EEG/LFP record as provided by the end-user. SDN then automates scoring of the remainder of the EEG/LFP record. A novel REM scoring correction procedure further enhanced accuracy. SDN reliably scores EEG and LFP data and retains sleep-wake architecture in wild-type mice, in sleep induced by the hypnotic zolpidem, in a mouse model of Alzheimer's disease and in a genetic knock-down study, when compared to manual scoring. SDN reduced manual scoring time to 1/12. Since SDN uses transfer learning on each independent recording, it is not biased by previously scored existing data sets. Thus, we find SDN performs well when used on signals altered by a drug, disease model or genetic modification.

Keywords: Alzheimer’s Disease Model; Automated Sleep-Wake scoring; Basic Sleep Research; Deep learning; Hypnotics; In Vivo Pharmacology; Transfer learning; Translational Sleep Research.

Figure 1.. Sleep-Deep-Net (SDN) scores mouse EEG/EMG signal with high reliability compared with human expert scores and retains the proportion of sleep-wake states.

A. F1 scores for wakefulness, NREM and REM reveal reliable scoring respective to manual scores by the expert scorer on which SDN was trained. B. Shows the proportion of wake, NREM and REM from 24-hour records when determined by manual scoring. C. Shows the proportion of wake, NREM and REM from 24-hour records when determined by SDN automated scoring. On a 24-hour record-by-record basis, the proportion of wakefulness (D), NREM sleep (E), and REM (F) correlate highly when determined by manual scoring vs SDN automated scoring. Data were from eleven 24-hour recordings from six mice.

Figure 2.. Sleep-Deep-Net (SDN) retains sleep architecture of manual scoring.

A. A representative multitaper time-frequency plot of EEG from a 24 hour record with aligned hypnograms drawn from manual scores and SDN automated scores. B. Time-weighted bouts of wakefulness, NREM and REM as determined from manual scores. C. Time-weighted bouts of wakefulness, NREM and REM as determined from ADN automated scores. D. Scatter plots with fitted least squares line reveal close relationship between time-weighted bouts for all three sleep-wake states on a record-by-record basis. Pearson’s correlation coefficients reveal statistically highly significant correlations between all three sleep-wake states comparing manual vs SDN automated scores. Data were from eleven 24-hour recordings from six mice.

Figure 3.. Sleep-Deep-Net (SDN) retains sleep architecture of hypnotic-drug (zolpidem) induced sleep compared with manual scoring.

A. A representative multitaper time-frequency plot of EEG from a 3-hour record in which 5 mg/kg zolpidem was delivered by intraperitoneal injection with aligned hypnograms drawn from manual scores and SDN automated scores. B. F1 scores reveal high reliability of drug induced NREM sleep and overall reliability, with reliability of REM in a similar range to WT records. The relative time spent in each sleep-wake state as determined by manual scores (C) vs SDN scores (D) shows similar values. N=4

Figure 4.. Sleep-Deep-Net (SDN) retains sleep architecture in local field potential (LFP) data from a mouse model of Alzheimer’s disease (APP/PS1) compared with manual scoring.

A. F1 scores between manual and SDN automated sleep scores were computed with LFP recordings performed on 5 APP/PS1 mice. Each sleep-wake state and all states reveals high reliability of NREM sleep, wake, and all states combined, with reliability of REM in a similar range to the WT case shown in Fig.1. The relative time spent in each sleep-wake state as determined by manual scores (B) was highly similar to SDN scores (C). N = 5.

Figure 5.. The significant increase in NREM delta power due to knocking down the expression of GABA_A alpha3 subunits in the thalamic reticular nucleus (Uygun et al., 2022) is reproduced when SDN scored the data.

Sleep-wake states were determined by manual scoring in panels A-C, from Uygun et al., 2022. A. Baseline time-frequency power dynamics reveal elevated delta (1.5 – 4 Hz) during NREM prior to rapid eye movement sleep (REM) transitions, during the whole 12 h light period. B. Following the knock-down of alpha3 containing GABA_A receptors in the thalamic reticular nucleus (α3KD), high delta in NREM prior to REM transitions was further increased. C. In contrast to their own BL levels (blue), α3KD mice (red) produced more delta power during NREM prior to REM transitions [t (5) = 2.14, p = 0.04]. Sleep-wake states determined by SDN scoring in panels D-F reproduced the effects found by manual scoring [ t (5) = 6.6, p = 0.0006]. Significance was tested using one-tailed paired t-tests. Thick lines indicate mean; envelopes indicate SEM. N=6.