Available via license: CC BY 4.0
Content may be subject to copyright.
npj | digital medicine Article
Published in partnership with Seoul National University Bundang Hospital
https://doi.org/10.1038/s41746-024-01115-7
Generalized sleep decoding with basal
ganglia signals in multiple movement
disorders
Check for updates
Zixiao Yin 1,2 ,HuilingYu
3, Tianshuo Yuan1,ClaySmyth
4,MdFahimAnjum
5,GuanyuZhu
1,2, Ruoyu Ma1,
Yichen Xu1,QiAn
1, Yifei Gan1, Timon Merk2,GuofanQin
1, Hutao Xie1,NingZhang
6, Chunxue Wang6,
Yin Jiang7,FangangMeng
7, Anchao Yang1, Wolf-Julian Neumann2, Philip Starr 8, Simon Little 5,
Luming Li3,9 & Jianguo Zhang 1,7,10
Sleep disturbances profoundly affect the quality of life in individuals with neurological disorders.
Closed-loop deep brain stimulation (DBS) holds promise for alleviating sleep symptoms, however, this
technique necessitates automated sleep stage decoding from intracranial signals. We leveraged
overnight data from 121 patients with movement disorders (Parkinson’s disease, Essential Tremor,
Dystonia, Essential Tremor, Huntington’s disease, and Tourette’s syndrome) in whom synchronized
polysomnograms and basal ganglia local field potentials were recorded, to develop a generalized,
multi-class, sleep specific decoder –BGOOSE. This generalized model achieved 85% average
accuracy across patients and across disease conditions, even in the presence of recordings from
different basal ganglia targets. Furthermore, we also investigated the role of electrocorticography on
decoding performances and proposed an optimal decoding map, which was shown to facilitate
channel selection for optimal model performances. BGOOSE emerges as a powerful tool for
generalized sleep decoding, offering exciting potentials for the precision stimulation delivery of DBS
and better management of sleep disturbances in movement disorders.
A considerable fraction of the world’spopulationisaffectedbysleepdis-
orders, which result in substantial welfare expenses1–3. Irregular sleep pat-
terns and related disorders are robust indicators of morbidity and mortality
across all causes4,5.Specifically, movement disorders patients are a popula-
tion strongly affected by sleep disorders6, which profoundly impact their
quality of life and potentially hasten the disease progression7.Deepbrain
stimulation (DBS) is a therapy widely used across multiple movement
disorders, effectively ameliorating motor symptoms while also contributing
to the improvement of sleep disturbances8,9. Importantly, the recent advent
of the adaptive closed-loop DBS10 offers the unprecedented promise of
enhancing sleep quality further through sleep stage-specific stimulation11,12.
However, this advancement poses the need for automatic decoding of a
patient’s sleep-wake cycle.
Typically, the classification of sleep stages necessitates intricate
laboratory-level polysomnography (PSG) monitoring. Although wear-
able devices like actigraphy13 and photoplethysmography14 have the
potential for sleep staging, their data is lower dimensional than neural
signals, resulting in limited classification accuracies15. Actigraphy-based
classification in populations with movement disorders will likely also be
challenging and require formal validation. Moreover, dependency on
1Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing, China. 2Movement Disorder and Neuromodulation Unit, Department of
Neurology, Charité—Campus Mitte, Charite—Universitatsmedizin Berlin, Chariteplatz 1, 10117 Berlin, Germany. 3National Engineering Research Center of
Neuromodulation, School of Aerospace Engineering, Tsinghua University, 100084 Beijing, China. 4Department of Bioengineering, University of California, San
Francisco, UCSF Byers Hall Box 2520, 1700 Fourth St Ste 203, San Francisco, CA 94143, USA. 5Department of Neurology, University of California, San Francisco,
1651 4th Street, San Francisco, CA 94158, USA. 6Department of Neuropsychiatry, Behavioral Neurology and Sleep Center, Beijing Tiantan Hospital, Capital
Medical University, Beijing, China. 7Department of Functional Neurosurgery, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China. 8Depart-
ment of Neurosurgery, University of California, San Francisco, Eighth Floor, 400 Parnassus Ave, SanFrancisco, CA 94143, USA. 9IDG/McGovern Institute for Brain
Research, Tsinghua University, 100084 Beijing, China. 10Beijing Key Laboratory of Neurostimulation, Beijing, China. e-mail: zixiao_yin@ccmu.edu.cn;
Simon.Little@ucsf.edu;zjguo73@126.com
npj Digital Medicine | (2024) 7:122 1
1234567890():,;
1234567890():,;
additional wearable devices will reduce utility, especially in elderly
populations.
Prior research, including our own16–19,hasattemptedtodecodesleep
stages based on basal ganglia local fieldpotentials (LFP s)re corded from DBS
electrodes. This approach has proven to be feasible and potentially reduces
the use case for additional wearable devices. However, existing research has
predominantly focused on individualized models, where training and pre-
diction are conducted using data from the same patient. This approach
requires sleep labeling for each new patient. In clinical practice, the acqui-
sition of laboratory-level sleep recordings for every DBS patient to establish
individualized sleep staging models would not be tractable or scaleable20.In
this study, we addressed this challenge by establishing the largest to date
synchronized basal ganglia LFP - PSG dataset in a cohort of 121 patients
with movement disorders. With this dataset, we trained a sleep decoding
model, termed BGOOSE (colloquially “big goose”), the Basal Ganglia
Oscillation-based Model for Sleep Stage Estimation, that could decode sleep
stages across individuals, disease entities, and even basal ganglia nucleus
targets. We also investigated the role of additional electrocorticography
(ECoG) in sleep decoding and established the projection between electrode
localization and sleep decoding accuracy. Finally, we validated our
classification model on two external datasets recorded at different post-
operative time points and using different DBS devices.
Results
Patient demographics and determination of the best decoder
We recorded 169 overnight (80,430.5 min in bed) synchronized poly-
somnogram and basal ganglia field potentials in 141 patients with move-
ment disorders who were treated with DBS. Sleep stage was determined both
manually and algorithmically21 based on standard rules22 every 30 s and only
epochs with consistent judgments between human experts and algorithms
were qualified for further analyses (on average nepochs = 621.9 per night).
After the exclusion of recordings with a low count of NREM or REM sleep
(< 5 min), 140 overnight recordings from 121 patients including Parkinson’s
disease-subthalamic nucleus (PD-STN, n= 20), PD-globus pallidus inter-
nus (PD-GPi, n=8), Dystonia-STN (n= 48), Dystonia-GPi (n=23),
Huntington’sdisease(HD)-GPi(n= 11), Tourette syndrome (TS)-GPi
(n= 4) and Essential tremor-ventral intermediate nucleus/caudal zona
incerta/posterior subthalamic area (ET-vim/CZi/PSA, n= 7) were analyzed
(Fig. 1A). Demographics, disease information, and sleep parameters of each
disease-target group are shown in Table 1. Previous sleep decoding studies
Parkinson’s Disease – subthalamic nucleus (Nsub = 20, Nnight = 23)
Parkinson’s Disease – globus pallidus internus (Nsub = 8, Nnight = 10)
Dystonia – subthalamic nucleus (Nsub = 48, Nnight = 50)
Dystonia – globus pallidus internus (Nsub = 23, Nnight = 27)
Huntington’s Disease – globus pallidus internus (Nsub = 11, Nnight = 18)
Tourette Syndrome – globus pallidus internus (Nsub = 4, Nnight = 5)
Essential Tremor – ventral intermediate nucleus/caudal zona incerta/
posterior subthalamic area (Nsub = 7, Nnight = 7)
PD-STN
PD-STN
PD-GPi
PD-GPi
DYS-STN
DYS-STN
DYS-GPi
DYS-GPi
HD-GPi
HD-GPi
TS-GPi
TS-GPi
ET-Vim/
ET-Vim/
cZI/PSA
cZI/PSA
Within- and Cross-subject decodin
g
Overall, Nsub = 121, Nnight = 140
NREM
REM
Awake
Awake NREM REM
Predicted Label
True Label
Sleep recordings Raw data Feature construction Sleep staging model
> SMOTE oversampling
> Standard Scaler (Z transform)
Nested cross validation
> Bayesian optimization
> Inner 4-fold cross validation
> Outer 4-fold (within subject) or
LOO (cross subject) cross validation
Model evaluation
> Accuracy for model performance
> SHAP for feature importance
Pipeline
i. Influential factors for decoding performance
ii. ECoG
electrode analysis
iii. Prediction
network mapping
Data collection Model development
Extension analysis External validation
Tiantan
Training
Testing
External validation
Tsinghua UCSF
114 temporarily
externalized subjects
LOO
12 subjects
with PD (PINS)
3 subjects with
PD (Medtronic)
Full model W
REM NREM
NREM
A
C
B
D
accuracies (%)
> Oscillatory
Band power,
Power ratio;
> Waveform
Skewness,
Kurtosis;
> Entropy
Higuchi fractal
dimension;
> Connectivity
Coherence
Fig. 1 | The training and prediction pipeline of the BGOOSE.BGOOSE is the
acronym for the basal ganglia oscillation-based model for sleep stage estimation.
ADiagram for synchronized basal ganglia and polysomnography recordings in a
cohort of 121 patients who underwent DBS surgery (adapted from Yin et al.32).
BPipeline for model development. See “Methods”for more details. CThe extension
analysis includes: (i) moderator analysis investigating factors that may influence
decoding accuracies; (ii) evaluating model performance after taking ECoG signals
into consideration; and (iii) prediction networking mapping analysis which aims at
establishing the projection between channel localization and sleep decoding accu-
racy (adapted from Merk et al.27). DThe performance of BGOOSE is validated in two
external datasets where basal ganglia local field potentials were recorded using
sensing-enable devices during sleep.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 2
employing different classifiers have obtained varied accuracies16–19.We
evaluated the performance of eight commonly used machine learning
classifiersinourPD-STNdataset.Itisshown that the LightGBM classifier, a
lightweight gradient boosting framework based on decision tree
algorithm23,24 that has been used in multiple sleep-related studies21,25,was
constantly associated with the highest accuracies as well as a sensible model
convergent speed in both the three-class (wake/NREM/REM) and five-class
(wake/N1/N2/N3/REM) decoding contexts (Supplementary Fig. 1). We
therefore based our subsequent analysis on the LightGBM classifier.
Individualized sleep decoding in different movement disorders
We then built individualized models to investigate whether sleep decoding
with basal ganglia signals was feasible in other disease entities than PD and
other basal ganglia nuclei than the STN. All subcortical recording sites are
showninFig.2A. Results showed that field potentials recorded from mul-
tiple basal ganglia nuclei including the STN, GPi, Vim, CZi, and PSA all
enabled satisfactory classifications of sleep-wake state in multiple disorders
including PD, DYS, HD and ET (Fig. 2B–I). The average accuracy was
94.1 ± 0.8% (range 93.7–95.7%) for the three-class classification and
86.5 ± 1.1% (range 84.7–87.6%) for the five-class classification across all
groups.
Cross-subject, cross-disease, cross-target, and across-all
decoding
Though individualized models demonstrated promising performances, they
require cumbersome PSG-based stage labeling and subcortical LFP
recordings for model training for each new patient, which could be chal-
lenging to obtain in the clinic. We explored whether sleep decoding could be
performed in a generalized manner, e.g., training a plug-and-play model
that could predict sleep stages in unseen patients. ET group was not analyzed
here due to their high heterogeneity in nucleus implantation (e.g., vim/CZi/
PSA). Data from the remaining 114 patients were analyzed. In the main
analysis, we focused on the three-stage classification of wake/NREM/REM
while a further attempt to classify NREM substages (i.e., NREM 1/2/3) was
performed as a side analysis. Through a rigorous leave-one-subject-out
cross-validation approach (the number of sleep stages used for training and
testing is provided in Supplementary Table 1), results (Fig. 3) showed that in
the same-disease, same-target, cross-subject manner, the average accuracy
of the generalized decoders was 83.9 ± 2.7%. For the same-disease, cross-
target decoding, the average generalized decoding accuracy was 85.1 ± 2.8%.
For the cross-disease, same-target decoding, the average generalized
decoding accuracy was 86.2 ± 0.5%. When not considering variabilities
from disease, target, and subject, the final average accuracy of the “cross-all”
models remained to be 85.9%. The average mislabeling rate for the cross-all
model was 8.85 ± 5.1 min per hour (Supplementary Table 2). For each stage,
the accuracies of the cross-all model in classifying wakefulness, NREM, and
REM sleep were 64.2%, 90.6%, and 71.8% respectively, with an overall
balanced accuracy of 77.6% (Supplementary Fig. 2) and a weighted F1 score
of 0.859 (Supplementary Fig. 3). Feature importance analysis with SHAP
(SHapley Additive exPlanations) showed that delta/theta power ratio was
the most contributed feature for the cross-all decodingfollowingbytheta
power and permutation entropy (Supplementary Fig. 4). For the further
classification of NREM 1/2/3 stages, the cross-all model obtained an average
accuracy of 78.2%, a balanced accuracy of 74.6% and a weighted F1 score of
0.806, with the identification of NREM2 sleep showing the highest average
accuracy of 80.2% (Fig. 4).
Additional electrocorticography helps sleep decoding
Accumulating evidence shows that electrocorticography (ECoG) adds
accuracy in informing state-dependent adaptive DBS26–28. Here we
investigated whether ECoG signals could outperform basal ganglia ones
in sleep decoding and whether a combination of both signals adds benefit
to each alone. In 8 temporarily externalized subjects (6 PD, 2 DYS), 11
overnight ECoG data were obtained together with basal ganglia and PSG
recordings. A total of 58 ECoG channels were analyzed (shown in Fig.
5A). We extracted ECoG features in frequency, time, and entropy
domain following the same manner as subcortical signals. Cortical-
subcortical connectivity was quantified as coherence (exemplified in Fig.
5B). Results showed that cortical and subcortical features resulted in
comparable sleep decoding accuracies, both significantly higher than that
obtained by connectivity features (Fig. 5C). A combination of subcortical
and cortical features significantly enhanced decoding accuracy compared
to each feature alone, especially for the cross-subject decoding (Fig. 5C),
with no further accuracy improvement after addin g connectivity features
Table 1 | Patient demographics and sleep parameters
All-
subjects
(n=121)
PD-
STN (n=20)
PD-GPi (n=8) DYS-
STN (n=48)
DYS-
GPi (n=23)
HD-
GPi (n=11)
TS-GPi (n=4) ETa(n=7)
Age 49.7 ± 16.9 62.1 ± 10.7 51.9 ± 21.0 47.7 ± 16.8 47.1 ± 16.2 47.1± 11.1 20.8 ± 2.2 56.1 ± 16.2
Sex (male) 49 (40.5%) 8 (40.0%) 6 (75.0%) 18 (37.5%) 9 (39.1%) 6 (54.5%) 1 (25.0%) 1 (14.3%)
BMI 22.4 ± 3.3 24.0 ± 3.4 21.4 ± 2.9 22.3 ± 3.5 22.0 ± 2.5 21.5 ± 1.5 24.9 ± 7.3 20.2 ± 2.5
Disease dura-
tion (year)
–9.8 ± 5.7 8.6 ± 5.6 7.1 ± 8.3 2.8 ± 1.9 6.1 ± 4.0 12.2 ± 3.9 12.0 ± 11.6
Disease severityb–40.2 ± 12.9 43.9 ± 18.1 21.3 ± 23.1 14.7 ± 11.8 63.2 ± 12.4 71.2 ± 5.6 36.1 ± 5.9
Sleep structure (%)c9.8/62.2/
10.1/17.8
11.3/62.3/
9.6/16.8
7.9/62.8/
12.9/16.4
10.9/60.2/
11.3/17.6
8.5/63.5/
8.3/19.7
8.3/68.7/
5.9/17.1
3.9/59.0/
13.2/23.9
11.2/62.6/
11.0/15.2
Sleep latency (min)d25.2 ± 31.2 38.2 ± 42.9 30.1± 37.9 26.6± 32.3 21.1 ± 21.6 14.5 ± 22.5 16.4 ± 11.5 9.7 ± 7.7
Sleep efficiency (%)e72.8 ± 14.6 70.3 ± 11.4 73.0 ± 16.5 73.3 ± 15.6 75.4 ± 12.8 69.8 ± 13.2 83.6 ± 6.5 65.8 ± 21.6
Sleep segmenta-
tion (n)f
12.6 ± 6.8 12.9 ± 5.5 11.2 ± 4.5 12.7 ± 7.6 10.0 ± 5.6 18.5 ± 7.8 10.2 ± 5.3 14.0 ± 6.2
PD Parkinson’s disease, STN subthalamic nucleus, GPi globus pallidus internus, DYS dystonia, HD Huntington’s disease, TS Tourette syndrome, ET essential tremor, BMI body mass index.
aTargets for ET included ventralis intermediate nucleus, caudal zona incerta, subthalamic nucleus, and posterior subthalamic area.
bThe presented disease severity scores were the MDS-Unified Parkinson’s Disease Rating Scale section III off-medication score for Parkinson’s disease, the Burke–Fahn–Marsden Dystonia Rating Scale-
Movement score for dystonia, the Unified Huntington’s Disease Rating Scale score for Huntington’s disease, the Yale Global Tic Severity Scale total score for Tourette syndrome, and the Essential Tremor
Rating Assessment Scale score for essential tremor.
cData presented represent the percentage of N1/N2/N3/REM sleep.
dSleep latency is defined as the time from lights out until the first epoch of any stage of sleep.
eSleep efficiency is defined as the ratio of total sleep time to time in bed.
fSleep segmentation is defined as the times that sleep is interrupted by > 2 min of wakefulness.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 3
(Fig. 5C). Subcortical and cortical features were equally important based
on SHAP analysis (Fig. 5D).
Moderator analysis and prediction of best decoding channels
We next investigated what factors may influence sleep decoding
accuracies. Linear mixed-effect models showed that patient demo-
graphics and disease/target information had no significant impact on
decoding performances in either individualized or generalized decodings
using basal ganglia and cortical features (Fig. 6A, left). Poorer sleep
quality as assessed by a higher count of sleep fragmentation was sig-
nificantly associated with lower decoding accuracies in basal ganglia-
based models (individualized model: coef = −0.00189, p= 1.01e–6;
generalized model: coef = −0.00317, p= 6.39e-4, Fig. 6A, right). A higher
proportion of N1 sleep was also associated with lower decoding
accuracies in cortical-based individualized models (coef = −0.00265,
p= 2.90e-4). We also evaluated the effects of various re-referencing
methods on the decoding performance of DBS channels. Specifically, we
compared the adjacent referencing method (e.g., the 1–2 re-referenced
channel for the 1-2-3-4 contact DBS lead) and the ‘sandwich’referencing
method (e.g., the 1–3 re-referenced channel). We found that the sand-
wich montage channels showed significantly higher average decoding
accuracies than the adjacent re-referenced channels in both the gen-
eralized and individualized models (Supplementary Fig. 5), a finding that
could be relevant to the application of adaptive DBS.
Fig. 2 | Individualized sleep decoding with basal ganglia signals for patients with
movement disorders. A shows lead (left) and contact (right) localizations for all
subjects. Bshows the true and predicted hypnograms from a representative patient
(DYS-STN #11). C–Ishows the lead localization and results of individualized sleep
decoding (training and testing the model using data from the same subject) in each
movement disorders cohort. The raincloud plot shows a cloud of individual raw data
points, a box plot, and a one-sided violin plot. For the boxplot, the lower and upper
borders of the box represent the 25th and 75th percentiles, respectively. The cen-
terline represents the median. The whiskers extend to the smallest and largest data
points that are not outliers (1.5 times the interquartile range). The violin plot shows
the probability density of the accuracies at different values. The portrait of the patient
with Parkinson’s disease is adapted from Arora et al.66.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 4
As previous works showed18,27, the anatomicallocation of the recording
electrode was also found to have an impact on decoding performance. This
relationship is important since for an unseen subject we are required to
determine the best possible decoding channel without having prior
knowledge. Within the framework of prediction network mapping27,we
simulated a volume of tissue recorded18 for each basal ganglia LFP channel
and calculated its connectivity with normative functional MRI
connectome29. We then correlated the connectivity strength with decoding
accuracy to establish the projection between channel localization and
decoding performance. This generated a whole-brain optimal decoding
map (Fig. 6B, left). Increased projections to the temporal lobe were asso-
ciated with higher decoding performances while increased projections to the
parietal lobe were associated with lower decoding accuracies, across all basal
ganglia seed targets. We validated this prediction network mapping
approach by constructing a map for all patients minus one and then using
the constructed map to predict the left-out unseen patients, repeating until
all patients were predicted. Results suggested that in this leave-one-subject-
out manner, higher spatial similarity (i.e., Spearman rho) between the basal
ganglia channel’s projection map and the optimal map constructed using
all-minus-one data significantly predicted higher decoding accuracies (coef
= 0.233, P= 5.50e-8, linear mixed effect model, Fig. 6B, right), supporting
the utility of the optimal map in predicting the best decoding channels in
unseen subjects based solely on channel localizations. Note that only adja-
cent but not sandwich re-referenced channels were employed here as
sandwich channels may have different sizes of volume recorded and signal-
to-noise ratios. We also attempted to construct an optimal decoding map for
ECoG channels. However, for ECoG this approach did not yield significant
predictions of decoding accuracy (Supplementary Fig. 6).
External validations using chronically embedded sensing-
enabled devices
In the final part, all patients’data (n= 114) were utilized to train a gen-
eralized sleep decoding model which we termed BGOOSE. Two external
datasets, the Tsinghua dataset, which included 12 PD patients with STN
LFPs recorded using the PINS sensing-enabled devices, and the UCSF
dataset, which included 3 PD patients with basal ganglia LFPs (2 STN, 1
GPi) recorded using the Medtronic sensing-enabled RC +S Summit
devices were used to test the performance of BGOOSE. Both datasets were
obtained at least one month after electrode implantation. Patient infor-
mationisshowninSupplementary Table 3. The average decoding
accuracies were 78.3 ± 6.1% and 79.8 ± 4.2% for the Tsinghua and UCSF
datasets, respectively, ranging between 86.7% and 66.7% for each individual
(Fig. 7A, B). For different stages, average accuracy was the highest for NREM
sleep: 81.7%, and was the lowest for REM stage: 56.3%, with an overall
Fig. 3 | Cross-subject sleep decoding with basal ganglia signals for patients with
movement disorders. Eleven decoding contexts are generated from different
combinations of 7 movement disorder cohorts. For the ALL-CROSSED condition, all
subject’s data (n= 114) were used for leave-one-subject-out cross-validation
(LOOCV) regardless of the disease and target differences. For the ALL-STN and
ALL-GPi conditions, data from patients who had implantations in the STN (n= 67)
and GPi (n= 47) were used for LOOCV, respectively, regardless of the disease
information. For the ALL-PD and ALL-DYS conditions, data from patients who were
diagnosed as PD (n= 28) and dystonia (n= 71) were used for LOOCV, respectively,
regardless of the DBS target information. For the PD-STN (n= 19), PD-GPi (n= 9),
DYS-STN (n= 48), DYS-GPi (n= 23), HD-GPi (n= 11), and TS-GPi (n= 4) condi-
tions, data from each disease-target group were used for LOOCV in their individual
groups. Decoding accuracies are shown in the middle column, with accuracy (ACC),
balanced accuracy (BA), and F1 score showing below the name of the cohort. Each
dot represents the accuracy value obtained from the left-out subject in that group.
The deep-colored regions in the one-sided violin plots show the probability density
of the decoding accuracies obtained from the best decoding channels. The light-
colored regions show the decoding accuracies obtained from all channels. The
vertical gray line represents the mean accuracy value of the best decoding channels.
Decoding accuracies for each stage of wakefulness, NREM, and REM sleep from the
best decoding channels are shown in the right column. The dashed black line
represents the chance accuracy of 33%. The portrait of the patient with Parkinson’s
disease is adapted from Arora et al.66.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 5
mislabeling rate of 12.8 ± 3.4 min per hour. A further subdivision of NREM
sleep into N1/N2/N3 stages demonstrated high accuracies in identifying
N2 sleep (90.7%), while the accuracies of classifying N1 (43.9%) and N3
(5.6%) stages were low (Supplementary Fig. 7). Decoding performances
across all 15 subjects in classifying wake/NREM/REM stages were sig-
nificantly predicted by the optimal decoding map (Fig. 7C). Channels
predicted to have good performances had significantly higher accuracies
than all-channel average accuracies, though still lower than the theoretically
best ones (Fig. 7D, E).
Discussion
We developed a novel sleep staging tool named BGOOSE (colloquially
“big goose”), designed for the automated decoding of sleep stages using
local field potentials recorded from DBS electrodes in movement dis-
orders. The development of BGOOSE was preceded by the establishment
of the largest known database of sleep-related basal ganglia electro-
physiological recordings, wherein precise sleep stage annotations were
applied through both manual and algorithmic methods. Subsequently,
our study demonstrated the feasibility of generalized sleep decoding,
achieving an average decoding accuracy of approximately 85% in a cross-
patient, cross-disease, and cross-basal ganglia structure scenario for the
classification of wake/NREM/REM stages. Moreover, we proved that the
inclusion of additional intracranial ECoG electrodes in the frontoparietal
cortex further enhanced sleep stage decoding accuracy, which might be
particularly important in patients with sleep fragmentation who showed
lower decoding accuracy. By mapping the relationship between electro de
contact locations and decoding accuracy, we provided an approach for
channel selection for optimal sleep decoding performance, using neu-
roimaging data only. Finally, the generalizability of our approach was
validated in two independent external validation datasets collected using
different DBS devices. A pipeline showing how BGOOSE could be opti-
mally used for sleep stage decoding is provided in Fig. 8.
Prior research on automatic sleep staging has predominantly been
based on EEG data or SEEG data30,31. The demand for sleep state decoding
using basal ganglia LFP data has arisen dramatically due to the recent
application of adaptive DBS which could potentially better address sleep
disturbances through sleep stage-specific stimulation10,32.Oneoftheearliest
models for automatic sleep staging using basal ganglia signals was proposed
by Thompson et al. in19, who trained a support vector machines (SVM)-
based model with STN LFPs for sleep decoding in 10 patients with PD. The
model exhibited promising performance in individualized scenarios,
achieving approximately 90% overall accuracy in classifying awake/NREM/
REM states, but fell short in cross-subject scenarios with an average accuracy
of about 50%. Subsequent studies16,17 explored the use of artificial neural
network (ANN), a type of deep learning technology, and decision tree
models for sleep staging in PD-STN patients, with similar performance for
individualized decoding and around 65% accuracy for cross-subject
decoding. Our previous work18 trained random forest models with over
500 h of pallidal LFP data recorded during sleep but still only obtained a
cross-subject decoding accuracy of 65.1%. Given that the patterns of basal
ganglia activity across sleep stages were different from patient to patient19
and that diseases were likely to further exert alterations on brain activity18,it
could be challenging to build a generalized model for all movement dis-
orders with acceptable accuracies. Here in this study, the BGOOSE model we
presented, based on the LightGB M gradient boosting framework23, achieved
approximately 85% sleep staging accuracy in a generalized context and
maintained around 80% accuracy duringout-of-cohortvalidations. Given
that the interrater reliability of sleep stage scoring for human experts is
around 80%33, we considered that the accuracies generated by BGOOSE
could be good enough to use even though we addressed only the three-stage
rather than the five-stageproblem.Intheexactusecaseofclosed-loopDBS,
since there are currently very few, if not no, proposed reasonable acceptance
criteria needed for effective closed-loop DBS during sleep, it is difficult to
interpret how good such accuracies are. In our previous proof-of-principle
Fig. 4 | Cross-subject decoding for the classification of NREN 1/2/3 stages with basal ganglia signals for patients with movement disorders. The same convention as in
Fig. 3.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 6
study12, a classifier model was trained to identify N3 sleep epochs using
ECoG signals to inform adaptive DBS during sleep. The decoder achieved
high specificity (0.94 ± 1.4e–2) and well above chance sensitivity
(0.62 ± 4e–2) with an overall accuracy of around 85%. With this staging
performance, the aDBS for sleep stage targeting can already have a sig-
nificant impact on sleep structures. Based on the mislabeling rate of
9–12 min per hour with BGOOSE presented here, it could be inferred that
approximately 6 h of sleep can be correctly labeled in a given night of 7-h
sleep. This should reasonably provide a remarkable change to the current
constant-parameter stimulation strategy during the night.
In the present study, 28 features were constructed for the training of the
BGOOSE model, including oscillatory features, waveform features, and
nonlinear entropy features. These features were selected based on previous
work in features-based automatic sleep staging algorithms using EEG or
LFP signals16,18,19,21,34. For example, it was shown that the permutation
entropy of the EEG21 and LFP spectral powers16 are among the most con-
tributed features for accurate sleep staging. We found that the LFP power
ratio between the delta and theta band was the most important feature for
cross-subject sleep decoding. This could be because delta and theta oscil-
lations are the defining features of NREM35 and REM36,37 sleep, respectively.
Therefore, it is sensible to speculate that a high, low, and medium delta/theta
power ratio may indicate NREM, REM, and awake stages, respectively,
although this is not necessarily applicable to all subjects, as in patients with
low decoding accuracies the rank of feature importance was very different
from that in the whole cohort (Supplementary Fig. 8). Importantly, the
features included in the current study were chosen to be computationally
efficient for the future online implementation. Other features, which are
more computationally intensive and have been demonstrated to differ-
entiate sleep stages including phase-amplitude coupling38 and oscillatory
bursts32 can also be included in future matrices with the advent of more
powerful embedded microprocessors.
An interesting finding in the moderator analysis was that higher
numbers of sleep fragmentation and a higher proportion of N1 sleep were
associated with lower decoding accuracy. This finding is in line with what
Vallat et al. found in their automated sleep decoding work with EEG signals,
where the percentage of N1 sleep and percentage of stage transitions were
the two top predictors of worse accuracy21. This relationship could be
plausible as the N1 stage is associated with the highest interrater variability
even for human experts33. However, in practical terms, this might imply that
the accuracy of automated scoring could diminish in patients with severe
sleep disorders, a population who particularly require the most sleep-stage-
specific intervention. Adding further sources of neural activitie s (e.g., ECoG)
would help improve decoding accuracies as shown in Fig. 4.
Another factor that could have an impact on decoding accuracy is the
re-reference approach for DBS channels.Traditionally,thefour-contact
DBS lead was re-referenced adjacently, creating three bipolar channels,
which is preferable if one wants to magnify local events that occur near the
contacts39. The advent of adaptive DBS introduced a “sandwich”montage,
which has the advantage that stimulation signals can be passively canceled
out through “common mode rejection”20. We found that sandwich-
referenced channels were associated with significantly higher average
decoding accuracy than the adjacent referenced channels even when sti-
mulation was not turned on. We speculated that this could be because the
1–3or2–4 referenced approach enables the generated channel to record a
larger area of neuronal activity, thus potentially enhancing the signal-to-
noise ratio and preventing the cancellation of symmetrically recorded local
sources. Given that the external dataset involved channels generated using
both reference approaches and the decoding accuracies could still be overall
CORTICAL POWER
CORTICAL-BASAL GANGLIA COHERENCE
HYPNOGRAM
Time (hours)
Stage Freq (Hz)Freq (Hz)
A
B
CD
***** *
**
** FEATURE IMPORTANCE
ns
DBS
ECoG
COH
Y
ns
*** ns
**
Fig. 5 | Additional electrocorticography (ECoG) electrode improves sleep
decoding. A shows the localization of all temporary ECoG electrodes in a glass brain.
Bshows an example time-frequency representation of cortical power (upper),
cortical-basal ganglia coherence (middle), and the corresponding hypnogram
(bottom) in a representative subject. Cshows the decoding accuracies using DBS
channel features, ECoG channel features, cortical-basal ganglia coherence features
(COHY), DBS plus ECoG channel features, and all features together (ALL) in the
contexts of within-subject decoding (left) and cross-subject decoding (right). The
gray dashed horizontal lines indicate a chance accuracy of 33%. *P< 0.05,
** P< 0.01. Dshows the feature importance when conducting cross-subject
decoding with all features. ECoG features are as important as basal ganglia features
(P= 0.993, Mann–Whitney Utest), though both features are more important than
coherence features, as shown in the inset. ns, non-significant.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 7
predicted by the optimal projection map, we believe that the montage dif-
ference would not remarkably influence the application of the optimal
projection map in guiding the selection of better decoding channels.
We would like to highlight several key features of BGOOSE.(1)The
most distinctive characteristic of BGOOSE lies in its plug-and-play, gen-
eralized decoding capability across various movement disorders. Unlike
previous research that often focused on single basal ganglia LFP signals for a
specific movement disorder, BGOOSE encompasses five of the most com-
mon movement disorders treated with DBS, including PD, DYS, HD, TS,
and ET, as well as the two most frequently used DBS targets, STN, and
GPi40–42. This breadth of applicability represents a significant clinical
potential with immediate relevance. (2) BGOOSE also stands out for its use
of single-channel data and exceptionally fast feature extraction and pre-
diction speed. In our testing, BGOOSE was able to construct features and
perform classification for a 30-s sleep epoch in less than 0.005 s on an Intel i5
consumer laptop. Although it must be noted that hyper-threaded processors
are not easily comparable to the single-threaded reduced instruction set
processor which aims to conserve size and power as would be found on an
implantable pulse generator (IPG), BGOOSE’s feature extraction process
that avoids computational complex indices is still advantageous for the
future implementation of low-latency online decoder within the IPG.
Moreover, the advantage of requiring only a single channel’sdataisparti-
cularly beneficial in the context of aDBS, as stimulation channels typically
lack recording functionality20,43.(3)BGOOSE also explores the integration of
ECoG signals, demonstrating the potential for enhanced decoding accuracy
when combined with cortical inputs. The chronically implanted ECoG
electrodes have been showing increasingly promising potential in assisting
state decoding for aDBS10,12,27,28. While the current version of BGOOSE
focuses on generalized models trained with basal ganglia signals due to
limited ECoG data, as more ECoG electrodes are implanted in patients with
movement disorders, future iterations of BGOOSE are expected to incor-
porate cortical features, further improving decoding performance. (4)
Lastly, BGOOSE’s provision of an optimal decoding map allows for the
identification of potential optimal basal ganglia decoding channels wit honly
lead localization information18,27. This feature becomes essential, especially
when dealing with new patients where ground truth sleep labels are not
available.
BGOOSE presents potential applications across several contexts.
Firstly, and most directly, it serves as a tool in guiding stage-specificsti-
mulation for closed-loop DBS systems12. Current closed-loop algorithms for
PD were predominantly informed by wakefulness beta power to adjust
stimulation parameters, but beta activity has been shown to vary sig-
nificantly across different sleep stages18,44. Failure to identify a patient’s
current sleep state and persistently using the wakeful beta activity level as a
threshold may result in either insufficient or excessive stimulation during
sleep32.WithBGOOSE, beta thresholds used for informing stimulation can
be adjusted in pace with sleep stage transitions (e.g., adjusting the beta
threshold to a lower level during NREM sleep so that pathological beta in PD
could be better suppressed through stimulation32,45). In addition, obtaining
the patient’s sleep stage enables the possibility of enhancing specific bene-
ficial waveforms relevant to different sleep stages, such as spindle and slow
wavesinNREMsleep
46 and sawtooth waves in REM sleep47.Secondly,for
patients whose symptoms are not present at night (e.g., essential tremor and
Tourette syndrome), being able to obtain sleep/awake classifications based
on LFP signals through BGOOSE means that we could switch off stimula-
tion automatically during sleep when stimulation is notneeded. Theconcept
T-value
**
**
**
Demographics Target/disease Sleep quality Sleep structure
P = 6.4e-04
DBS-CROSS (Lateral) DBS-CROSS (Medial)
A. Moderator analysis on the decoding performance
B. Prediction network mapping for DBS channels
P = 5.5e-08
Coefficient (z)
DBS-INDIV
(Nsub=121)
DBS-CROSS
(Nsub=114)
ECoG-INDIV
(Nsub=8)
ECoG-CROSS
(Nsub=8)
Fig. 6 | Moderator analysis and network mapping of the decoding accuracies.
AHeatmap showing the influential factors of decoding accuracies in different
contexts including within-subject decoding with basal ganglia data (DBS-INDIV),
cross-subject decoding with basal ganglia data (DBS-CROSS), within-subject
decoding with ECoG data (ECoG-INDIV), and cross-subject decoding with ECoG
data (ECoG-CROSS). Candidate factors are elaborated on in the Methods section.
The right side shows the regression plot which depicts the correlation between the
number of sleep fragmentations and the decoding performances. Significant asso-
ciations with pvalues < 0.0029 (0.05/17) were highlighted with asterisks. Bshows the
lateral and medial view of the optimal decoding map for the basal ganglia electrodes.
The map was generated by first generating a volume of tissue recorded with a radius
of 5 mm for each contact and then calculating the connectivity pattern between the
seed volume and the normative functional MRI connectome. The obtained whole-
brain connectivity strength was then voxel-wised correlated with the decoding
accuracy, resulting in the final optimal decoding map. Increased projection to the
purple area indicates a higher chance to obtain good decoding results while increased
projection to the blue area indicates a lower chance to obtain good decoding results.
The right panel shows the repeated measurement regression plot between the spatial
similarity to the optimal map and the decoding accuracies obtained in a leave-one-
subject-out manner.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 8
of “deep brain stimulation holidays”in essential tremor48 has long been
proposed, where stimulation is temporarily stopped during the night to
prevent tolerance to DBS. Given that the adherence of manually switching
on & off stimulation might be difficult for patients49,BGOOSE emerges as a
useful tool to realize this aim algorithmically. Thirdly, the application of
BGOOSE can facilitate the sleep management in patients with movement
disorders. Using the DBS device for sleep staging potentially eliminates the
need for additional wearable devices, offering valuable 24/7 longitudinal
data for drug adjustments and DBS programming11. Fourthly, the use of
BGOOSE has the potential to advance sleep research in movement disorder
Confusion matrices for the 15 external subjects
Tsinghua#10
A
B
CDE
**** ***
Coef = 0.253, P= 1.2e-05
True label
Predicted label
(optimal map)
(individual projection)
Fig. 7 | External validation for the BGOOSE. Ashows the average accuracy for the
BGOOSE to classify sleep stages in two external cohorts with basal gangli a recordings
during sleep. The lower and upper borders of the box represent the 25th and 75th
percentiles, respectively. The centerline represents the median. The whiskers extend
to the smallest and largest data points that are not outliers (1.5 times the interquartile
range). The black dashed line represents the chance accuracy of 33%. Bshows the
decoding confusion matrices for each of the 15 unseen subjects in the two cohorts.
Cdemonstrates the repeated measurement correlation between spatial similarity to
the optimal map and the decoding accuracy. The upper inset shows the medial,
posterior, and dorsal views of the optimal decoding map. Dshows the correlation
between spatial similarity and decoding accuracy in one representative patient
(Ts#10). The channel with higher decoding accuracy (upper inset) has higher whole-
brain projection similarity to the optimal map than the channel with lower decoding
accuracy (bottom inset). Eshows the comparison between mean accuracy values
obtained from optimal-map selected channels (Opt.map), channels with worst
decoding performances (Min), all channels (Avg), “sandwich”referenced channels
(Sand), and channels with best decoding performances (Max). Same conventions as
in Fig. 7A. P= 9.82×10−4for the comparison of accuracy between map-based
channels and worst decoding channels. P= 2.01×10−3for the comparison of accu-
racy between map-based channels and all channels. P= 3.02×10−2for the compar-
ison of accuracy between map-based channels and sandwich re-referenced ch annels.
P= 7.69×10−3for the comparison of accuracy between map-based channels and best
decoding channels. Wilcoxon signed-rank test.
Fig. 8 | Pipeline for the optimal use of BGOOSE. A pipeline summarizing the main function of BGOOSE and how can it be optimally used.
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 9
patients as one of the current barriers of conducting sleep research lies in the
cumbersome determination of sleep stages, typically requiring complex PSG
monitoring at sleep laboratories50.WhileBGOOSE is not presented as a PSG
replacement, it can be viewed as an alternative in scenarios where con-
ducting PSG monitoring is challenging due to equipment or environmental
constraints, or whenpatient cooperation is limited. BGOOSE,basedonbasal
ganglia signals, can provide sleep labels in these situations, thereby advan-
cing sleep research in patients with movement disorders.
Despite its many advantages, BGOOSE has several limitations that
warrant future improvement. First, in cross-decoding, while it achieved
an accuracy of over 90% for NREM sleep stages, the classification
accuracies for the wake and REM stages were approximately 60–70%.
This could be because, compared to the more homogenous NREM sleep,
the variability among patients in wake and REM stages was larger32,51.
This can also be supported by the feature importance analysis shown in
Supplementary Fig. 8, where patients with low accuracies in cross-subject
decoding had very different top-contributed features from the whole
cohort (Supplementary Fig. 4). However, it was essential to note that our
previous research had suggested that pathological oscillations in NREM
sleep were highly relevant to sleep disorders in PD, making this a critical
sleep stage to target for higher accuracy decoding32. Additionally, as
NREM sleep constitutes over 70% of total sleep time, achieving robust
identification of NREM sleep remains critical for sleep decoding models.
Second, it is worth noting that REM classification was notably poor in the
second external dataset (UCSF) in patients recorded chronically using a
fully embedded sensing-enabled DBS (Summit RC +S) pacemaker.
Differences in e.g., line noise frequency, or frequency-specific normal-
ization effects may make the challenging REM versus NREM distinction
difficult to generalize across recording systems. Third, in the generalized
scenario, we put our focus on the three-class decoding problem.
Although we tried to build decoders to further classify NREM sleep into
N1/N2/N3 stages as a side analysis, the performance was not that satis-
factory, especially for the N3 stage (Supplementary Fig. 7). One potential
reason could be that the reduced proportion of N3 sleep due to sleep
disturbances in a considerable portion of movement disorder patients
complicates training and generalization. Accumulating more deep sleep
data from chronic recordings for model training45 or identifying N3 sleep
using ECoG electrodes12 are potential solutions. Fourth, BGOOSE relies
on feature extraction and tree-based algorithms for sleep staging recog-
nition, whereas deep learning algorithms have shown significant
potential in the sleep classification field31,52,53. Although they may require
longer training and parameter tuning time, exploring state-of-the-art
deep learning algorithms is a worthwhile endeavor. Fifth, since BGOOSE
is trained using data recorded during externalized periods, it may lack
resistance to electrocardiogram (ECG) artifacts. ECG artifacts are com-
mon in perceptive data54 and have been observed in the Tsinghua and
USCF datasets. In addition, the “micro-lesion effect”during the exter-
nalized period could also negatively influence the generalized ability of
BGOOSE as this effect could temporarily change the osci llatory pattern in
basal ganglia (e.g., beta power in Parkinson’s disease)55,56. Including more
chronically recorded data with all types of environmental artifacts in
training will enhance BGOOSE’s ability to generalize. Sixth, BGOOSE is
currently trained in the off-stimulation condition, and its performance
during on-stimulation states for sleep staging remains unknown, which is
a crucial consideration for continuous monitoring and stimulation45.
Seventh, BGOOSE is trained solely for nighttime sleep, and its applic-
ability for accurate daytime napping sleep staging is uncertain. Lastly,
BGOOSE serves as a tool for automatic sleep staging but the question of
how to implement adaptive stimulation following sleep staging remains a
critical, unresolved challenge. For instance, the optimal stimulation
approach in N1 to induce deep sleep, in N2/3 to enhance beneficiary
waves, and in REM sleep to potentially suppress REM sleep behavior
disorder are all areas requiring further investigation. Addressing these
questions holds the key to realizing the potential of sleep interventi on as a
therapeutic tool for managing various disorders.
Methods
Patients and surgery
A total of 141 patients with movement disorders scheduled to undergo DBS
surgery at Beijing Tiantan Hospital were enrolled in the study after
obtaining written informed consent. This study is approved by the IRB of
Beijing Tiantan Hospital and performed per the Declaration of Helsinki.
Inclusion criteria comprised the following: (1) well-defined disease diag-
nosis; (2) ability to cooperate with whole-night PSG recordings; and (3)
absence of structural lesions observed on MRI scans. Under the guidance of
a stereotactic frame system, four-contact DBS electrodes were implanted
into the predefined basal ganglia regions as per standard protocols57.The
accuracy of electrode placement was confirmed through intraoperative
electrophysiological recordings, temporary stimulation, and postoperative
anatomical computed tomography (CT) scans. In six patients with PD and
two patients with dystonia, an additional eight-contact ECoG electrode was
implanted into the right motor cortex region through the same burr hole for
DBS26. Following electrode implantation, patients returned to the ward for
sleep recordings, which were typically conducted within 3–7dayspost-
surgery.
Sleep recordings and staging
During lead externalization, patients underwent sleep recordings for 1–2
consecutive nights following previous routines32. Signal recording was
conducted using a JE-212 amplifier (Nihon Kohden, Tokyo, Japan) with a
sampling rate of 1000/2000 Hz. All drugs that could potentially influence
brain state were stopped throughout the recording session, including anti-
parkinsonism and anti-dystonia medications, and sleeping pills. Sleep PSG
labels were manually scored every 30 s according to the rules outlined in the
AASM manual version 2.6. Four sleep parameters were measured and
reported in Table 1: the sleep structure, defined as the percentage of N1/N2/
N3/REM sleep; the sleep latency, defined as the time from lights out until the
first epoch of any stage of sleep; the sleep efficiency, defined as the ratio of
total sleep time to time in bed; and the sleep segmentation, defined as the
times that sleep is interrupted by over 2 min of wakefulness32.Inaddition,to
enhance staging efficiency and reduce the impact of scorer subjectivity on
labeling results, this process was aided by an established open-source sleep
staging algorithm (https://github.com/raphaelvallat/yasa)21,whichwas
trained on over 30,000 h of PSG data. Only sleep epochs with consistent
manual and algorithmic judgments were qualified for further sleep decoding
analyses. On average, 621.9 sleep epochs per night were included in the
analysis. Sleep stages were categorized into either three (awake/NREM/
REM) or five categories (awake/N1/N2/N3/REM) in later analysis. Data
from each stage needed to be collected for at least 5 min to be included in the
analysis.
ECoG and LFP preprocessing and feature extraction
All signals were Butterworth notch filtered to reject the 50 Hz ambient noise
and harmonics, followed by downsampling to 200 Hz. Bipolar re-
referencing was applied to all adjacent contacts for ECoG and DBS elec-
trodes. Besides, for DBS electrodes, a “sandwich”re-referencing (e.g., 1–3
and 2–4 referencing for the 1-2-3-4 lead) was also applied to mimic the
clinical use case of adaptive DBS20. The sandwich-referenced channels were
not analyzed in the network mapping section (see below) as they might have
different shapes of recording field and signal-to-noise ratios from adjacent-
referenced channels. For feature extraction, we computed a total of 28 local
features, which included 18 frequency-domain features: one total absolute
power, seven relative energy features in different frequency bands (delta
1–4Hz, theta 4–8Hz, alpha 8–13 Hz, low beta 13–20 Hz, high beta
20–30 Hz, low gamma 30–45 Hz, high gamma 55–90 Hz), and ten energy
ratio features across frequency bands (delta, theta, alpha, beta, gamma
[4 +3+2+1]). Additionally, we computed 10 time-domain features,
consisting of seven statistical features, including standard deviation, inter-
quartile range, skewness, kurtosis, number of zero crossings, Hjorth
mobility, and Hjorth complexity, and three entropy features, including
permutation entropy, Higuchi fractal dimension, and Petrosian fractal
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 10
dimension21. Furthermore, for patients withECoG implants, in addition to
the abovementioned local features, we extracted 18 basal ganglia-cortical
connectivity features, comprising coherence in seven frequency bands
(delta,theta,alpha,lowbeta,highbeta,lowgamma,andhighgamma),ten
coherence ratio features across five frequency bands (delta, theta, alpha,
beta, gamma band [4 +3+2+1]), and one peak coherence frequency,
defined as the frequency at which the coherence peak occurred. Conse-
quently, in scenarios using single-site signals, 28 features were generated,
whereas in cases utilizing dual-site signals, 74 features were com-
puted (28*2+18).
Machine learning models
We tested eight machine learning models in our study, namely, ridge
regression, SVM, k-nearest neighbors (KNN), decision tree, random forest,
XGBoost, LightGBM, and artificial neural network. Ridge regression, SVM,
KNN, decision tree, and random forest were implemented using machine
learning algorithms available in the scikit-learn library58. XGBoost was
implemented using the code provided at https://github.com/dmlc/xgboost/
tree/master59. LightGBM was implemented using the code available at
https://github.com/microsoft/LightGBM/tree/master23.Theartificial neural
network was developed based on the TensorFlow framework60.Detailed
hyperparameter tuning for each model can be found in Supplementary
Table 4. Any unspecified parameters were set to their default values.
Model training, evaluation, and features importance
Model performance was evaluated using nested cross-validation, with both
outer and inner cross-validation set to 4 folds. A Bayesian Optimization
hyperparameter search, comprising 50 rounds, was employed for hyper-
parameter tuning using a balanced accuracy as the scoring function. Given
the imbalance in the distribution of sleep stage labels, the SMOTE method
from the imbalanced-learn library61 was utilized to oversample the minority
class labels. For the sake of simplicity, model performance was assessed
using average accuracy in the main text. Results with balanced accuracy and
weighted F1 score were shown when indicated. The speed of model fitting
was evaluated as 1/log (fitting time). Feature importance was assessed and
compared using the SHAP method, available at https://github.com/shap/
shap/tree/master62.
Moderator analysis
In the moderator analysis, we investigated factors that may influence the
accuracy of sleep stage decoding. The independent variables encompassed
four categories, namely, (1) demographic information, including sex, age,
and BMI), (2) disease and target information (e.g., disease: PD /DYS/HD/TS,
target: STN/GPi), (3) sleep quality information, including the number of
sleep fragmentations, sleep efficiency, total sleep time, and sleep latency, and
(4) sleep structure information including the proportion of time spent in
each sleep stage. The effects of different referencing approaches on decoding
performance were also investigated.
Localization of DBS and ECoG electrode
We reconstructed the DBS electrodes using the advanced electrode locali-
zation pipeline with default settings in Lead-DBS version 2.5.3 (MATLAB
2019b)63. ECoG electrodes were reconstructed using the established method
with FreeSurfer64. The positions of DBS and cortical electrodes were stan-
dardized to the MNI template (ICBM 2009b Nonlinear Asymmetric) for
group-level analysis.
Network mapping for choosing the best decoding channel
We implemented a prediction network mapping approach that was pre-
viously established18,27 to predict potential optimal decoding channels. This
method generates a simulated recording field for each recording site (radius
= 5 mm) and calculates its functional connectivity with the whole-brain
normative connectome (Brain Genomics Superstruct Project-10029). By
correlating the decoding accuracy of individual channels with the whole-
brain connectivity pattern projected from the corresponding recording field,
and incorporating significant moderators identified in the moderator ana-
lysis as covariates, we generate an optimal projection map featured in the
AAL3 atlas with a total number of parcellations of 16665.Wethencalculated
the spatial similarity between each channel’s projection map and the opti-
mal projection map (i.e., the Spearman correlation between two 166-length
one-dimensional arrays). For an unseen recording channel, the greater the
similarity between its functional connectivity pattern with the whole brain
and the optimal map, the higher the likelihood it may obtain higher
decoding accuracy.
External validation
We conducted external validation of our model using two separate datasets.
The first dataset, referred to as the Tsinghua dataset, comprised 12 PD
patients who underwent STN-DBS surgery16. The second dataset, the UCSF
dataset, included two PD patients who received STN-DBS and one PD
patient who received GPi-DBS45. All 15 patients underwent overnight
synchronous PSG and basal ganglia signal recording in the off-medication &
off-stimulation state. Based on the MNI coordinates, no significant targeting
differences were found between the validation group and the training
cohort. Demographic information of the external subjects can be seen in
Supplementary Table 3. More detailed demographics were reported in two
original papers: the 12 PD patients in the Tsinghua dataset are the same
patients named No.1 to No.12 in Table 1in Yue Chen et al.16,andthe3PD
patients in the UCSF dataset are the same patients named as PD2, PD7, and
PD9 in Table 1in Md Fahim Anjum et al.45. The validation process
encompassed two aspects: (1) validation of decoding performances of the
BGOOSE and (2) validation of the prediction of the optimal decoding
channel using the optimal projection map.
Statistical analysis
We aimed to conduct statistical comparisons using non-parametric tests
including the Mann–Whitney U test, Wilcoxon signed-rank test, and
Spearman correlation wherever possible. We employed linear mixed- effects
models for the moderator analysis. The significance threshold for two-sided
P-values was set at 0.05.
Reporting summary
Further information on research design is available in the Nature Research
Reporting Summary linked to this article.
Data availability
The original data are not yet openly available, as it is being used in ongoing
projects. We welcome enquires for sharing this as part of a collaboration,
please contact the corresponding authors.
Code availability
Thecodeforanalyzingthedataandthefinal version of the generalized
decoding model is made freely available at https://github.com/zixiao-yin/
BGOOSE. The sleep decoding models for the three and five-stage classifi-
cations can be found at https://osf.io/mt72e/ (https://doi.org/10.17605/OSF.
IO/MT72E).
Received: 23 December 2023; Accepted: 23 April 2024;
References
1. da Silva, A. A. et al. Sleep duration and mortality in the elderly: A
systematic review with meta-analysis. BMJ Open 6,
e008119 (2016).
2. Devore, E. E., Grodstein, F. & Schernhammer, E. S. Sleep duration in
relation to cognitive function among older adults: A systematic review
of observational studies. Neuroepidemiology 46,57–78 (2016).
3. Tobaldini, E. et al. Short sleep duration and cardiometabolic risk:
From pathophysiology to clinical evidence. Nat. Rev. Cardiol. 16,
213–224 (2019).
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 11
4. Chattu, V. K. et al. The global problem of insufficient sleep and its
serious public health implications. Healthc. (Basel) 7, 1 (2018).
5. Garbarino, S., Lanteri, P., Durando, P., Magnavita, N. & Sannita, W. G.
Co-morbidity, mortality, quality of life and the healthcare/welfare/
social costs of disordered sleep: A rapid review. Int. J. Environ. Res
Public Health 13, 831 (2016).
6. Bailey, G. A. et al. Sleep disturbance in movement disorders: insights,
treatments and challenges. J. Neurol. Neurosurg. Psychiatry 92,
723–736 (2021).
7. Schreiner, S. J. et al. Slow-wave sleep and motor progression in
Parkinson disease. Ann. Neurol. 85, 765–770 (2019).
8. Hasegawa, H. et al. The subcortical belly of sleep: New possibilities in
neuromodulation of basal ganglia? Sleep. Med Rev. 52, 101317 (2020).
9. Yin, Z. et al. A quantitative analysis of the effect of bilateral
subthalamic nucleus-deep brain stimulation on subjective and
objective sleep parameters in Parkinson’s disease. Sleep Med.
https://doi.org/10.1016/j.sleep.2020.10.021 (2020).
10. Gilron, R. et al. Sleep-aware adaptive deep brain stimulation control:
Chronic use at home with dual independent linear discriminate
detectors. Front Neurosci. 15, 732499 (2021).
11. Fleming, J. E. et al. Embedding digital chronotherapy into
bioelectronic medicines. iScience 25, 104028 (2022).
12. Smyth, C. et al. Adaptive deep brain stimulation for sleep stage
targeting in Parkinson’s disease. Brain Stimul. 16, 1292–1296 (2023).
13. Boe, A. J. et al. Automating sleep stage classification using wireless,
wearable sensors. NPJ Digit Med. 2, 131 (2019).
14. Radha, M. et al. A deep transfer learning approach for wearable sleep
stage classification with photoplethysmography. NPJ Digit Med 4,
135 (2021).
15. Fekedulegn, D. et al. Actigraphy-based assessment of sleep
parameters. Ann. Work Expo. Health 64, 350–367 (2020).
16. Chen, Y. et al. Automatic sleep stage classification based on
subthalamic local field potentials. IEEE Trans. Neural Syst. Rehabil.
Eng. 27, 118–128 (2019).
17. Christensen, E., Abosch, A., Thompson, J. A. & Zylberberg, J. Inferring
sleep stage from local field potentials recorded in the subthalamic
nucleus of Parkinson’s patients. J. Sleep. Res 28, e12806 (2019).
18. Yin, Z. et al. Pallidal activities during sleep and sleep decoding in
dystonia, Huntington’s, and Parkinson’s disease. Neurobiol. Dis. 182,
106143 (2023).
19. Thompson, J. A. et al. Sleep patterns in Parkinson’s disease: Direct
recordings from the subthalamic nucleus. J. Neurol. Neurosurg.
Psychiatry 89,95–104 (2018).
20. Neumann, W., Gilron, R., Little, S. & Tinkhauser, G. Adaptive deep
brain stimulation: From experimental evidence toward practical
implementation. Mov. Disord. mds.29415 (2023). https://doi.org/10.
1002/mds.29415.
21. Vallat, R. & Walker, M. P. An open-source, high-performance tool for
automated sleep staging. Elife 10, e70092 (2021).
22. Berry, R. B. et al. AASM Scoring Manual Updates for 2017 (Version
2.4). J. Clin. Sleep. Med.13, 665–666 (2017).
23. Ke, G. et al. LightGBM: A Highly Efficient Gradient Boosting Decision
Tree. in Proceedings of the 31st International Conference on Neural
Information Processing Systems 3149–3157 (Curran Associates
Inc., 2017).
24. Bentéjac, C., Csörgő, A. & Martínez-Muñoz, G. A comparative
analysis of gradient boosting algorithms. Artif. Intell. Rev. 54,
1937–1967 (2021).
25. Kim, W.-P. et al. Machine learning-based prediction of attention-
deficit/hyperactivity disorder and sleep problems with wearable data
in children. JAMA Netw. Open 6, e233502 (2023).
26. Yin, Z. et al. Cortical phase-amplitude coupling is key to the
occurrence and treatment of freezing of gait. Brain awac121 (2022).
https://doi.org/10.1093/brain/awac121.
27. Merk, T. et al. Electrocorticography is superior to subthalamic local
field potentials for movement decoding in Parkinson’s disease. Elife
11, e75126 (2022).
28. Gilron, R. et al. Long-term wireless streaming of neural recordings for
circuit discovery and adaptive stimulation in individuals with
Parkinson’s disease. Nat. Biotechnol. https://doi.org/10.1038/
s41587-021-00897-5 (2021).
29. Buckner, R. L., Roffman, J. L. & Smoller, J. W. Brain genomics
superstruct project (GSP). https://doi.org/10.7910/DVN/
25833 (2014).
30. Faust, O., Razaghi, H., Barika, R., Ciaccio, E. J. & Acharya, U. R. A
review of automated sleep stage scoring based on physiological
signals for the new millennia. Comput Methods Prog. Biomed. 176,
81–91 (2019).
31. Fiorillo, L. et al. Automated sleep scoring: A review of the latest
approaches. Sleep. Med. Rev. 48, 101204 (2019).
32. Yin, Z. et al. Pathological pallidal beta activity in Parkinson’s disease is
sustained during sleep and associated with sleep disturbance. Nat.
Commun. 14, 5434 (2023).
33. Lee, Y. J., Lee, J. Y., Cho, J. H. & Choi, J. H. Interrater reliability of sleep
stage scoring: a meta-analysis. J. Clin. Sleep. Med.18, 193–202 (2022).
34. Sun, H. et al. Large-scale automated sleep staging. Sleep 40,
zsx139 (2017).
35. Hubbard, J. et al. Rapid fast-delta decay following prolonged
wakefulness marks a phase of wake-inertia in NREM sleep. Nat.
Commun. 11, 3130 (2020).
36. Cowdin, N., Kobayashi, I. & Mellman, T. A. Theta frequency activity
during rapid eye movement (REM) sleep is greater in people with
resilience versus PTSD. Exp. Brain Res 232, 1479–1485 (2014).
37. Hong, J., Lozano, D. E., Beier, K. T., Chung, S. & Weber, F. Prefrontal
cortical regulation of REM sleep. Nat. Neurosci. 26, 1820–1832 (2023).
38. Amiri, M., Frauscher, B. & Gotman, J. Phase-amplitude coupling is
elevated in deep sleep and in the onset zone of focal epileptic
seizures. Front Hum. Neurosci. 10, 387 (2016).
39. Shirhatti, V., Borthakur, A. & Ray, S. Effect of reference scheme on
power and phase of the local field potential. Neural Comput 28,
882–913 (2016).
40. Johnson, M. D., Miocinovic, S., McIntyre, C. C. & Vitek, J. L.
Mechanisms and targets of deep brain stimulation in movement
disorders. Neurotherapeutics 5, 294–308 (2008).
41. Sui, Y. et al. Deep brain stimulation initiative: Toward innovative
technology, new disease indications, and approaches to current and
future clinical challenges in neuromodulation therapy. Front Neurol.
11, 597451 (2020).
42. Yin, Z. et al. An individual patient analysis of the efficacy of using GPi-
DBS to treat Huntington’sdisease.Brain Stimul. 13,1722–1731 (2020).
43. Little, S. & Brown, P. Debugging adaptive deep brain stimulation for
Parkinson’s disease. Mov. Disord. mds.27996 (2020). https://doi.org/
10.1002/mds.27996.
44. Zahed, H. et al. The neurophysiology of sleep in Parkinson’s disease.
Mov. Disord. 36, 1526–1542 (2021).
45. Anjum, M. F. et al. Multi-night naturalistic cortico-basal recordings
reveal mechanisms of NREM slow wave suppression and
spontaneous awakenings in Parkinson’s disease. bioRxiv (2023)
https://doi.org/10.1101/2023.06.23.546302.
46. Geva-Sagiv, M. et al. Augmenting hippocampal-prefrontal neuronal
synchrony during sleep enhances memory consolidation in humans.
Nat. Neurosci. 26, 1100–1110 (2023).
47. Frauscher, B. et al. Rapid eye movement sleep sawtooth waves are
associated with widespread cortical activations. J. Neurosci. 40,
8900–8912 (2020).
48. Garcia Ruiz, P., Muñiz de Igneson, J., Lopez Ferro, O., Martin, C. &
Magariños Ascone, C. Deep brain stimulation holidays in essential
tremor. J. Neurol. 248, 725–726 (2001).
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 12
49. Barbe, M. T. et al. Deep brain stimulation in the nucleus ventralis
intermedius in patients with essential tremor: habituation of tremor
suppression. J. Neurol. 258, 434–439 (2011).
50. Fiorillo, L. et al. U-Sleep’s resilience to AASM guidelines. NPJ Digit
Med 6, 33 (2023).
51. Jaggard, J. B., Wang, G. X. & Mourrain, P. Non-REM and REM/
paradoxical sleep dynamics across phylogeny. Curr. Opin. Neurobiol.
71,44–51 (2021).
52. Supratak, A., Dong, H., Wu, C. & Guo, Y. DeepSleepNet: A model for
automatic sleep stage scoring based on raw single-channel EEG.
IEEE Trans. Neural Syst. Rehabil. Eng. 25, 1998–2008 (2017).
53. Perslev, M. et al. U-Sleep: resilient high-frequency sleep staging. NPJ
Digit Med 4, 72 (2021).
54. Neumann, W.-J. et al. The sensitivity of ECG contamination to surgical
implantation site in brain computer interfaces. Brain Stimul. 14,
1301–1306 (2021).
55. Chen, C. C. et al. Intra-operative recordings of local field potentials
can help localize the subthalamic nucleus in Parkinson’s disease
surgery. Exp. Neurol. 198, 214–221 (2006).
56. Yin, Z. et al. Local field potentials in Parkinson’s disease: A frequency-
based review. Neurobiol. Dis. 155, 105372 (2021).
57. Yin, Z. et al. Balance response to levodopa predicts balance
improvement after bilateral subthalamic nucleus deep brain
stimulation in Parkinson’s disease. NPJ Parkinsons Dis. 7, 47 (2021).
58. Pedregosa, F. et al. Scikit-learn: Machine learning in python. J. Mach.
Learn. Res. 12, 2825–2830 (2011).
59. Chen, T. & Guestrin, C. XGBoost: A scalable tree boosting system. in
Proceedings of the 22nd ACM SIGKDD International Conference on
Knowledge Discovery and Data Mining 785–794 (ACM, 2016). https://
doi.org/10.1145/2939672.2939785.
60. Abadi, M. et al. Tensorflow: A system for large-scale machine learning.
in 12th {USENIX} Symposium on Operating Systems Design and
Implementation ({OSDI} 16) 265–283 (2016).
61. Lemaître, G., Nogueira, F. & Aridas, C. K. Imbalanced-learn: A python
toolbox to tackle the curse of imbalanced datasets in machine
learning. J. Mach. Learn. Res. 18,1–5 (2017).
62. Lundberg, S. M. & Lee, S.-I. A unified approach to interpreting model
predictions. in Advances in Neural Information Processing Systems
(eds. Guyon, I. et al.) vol. 30 (Curran Associates, Inc., 2017).
63. Horn, A. et al. Lead-DBS v2: Towards a comprehensive pipeline for
deep brain stimulation imaging. Neuroimage 184, 293–316 (2019).
64. Hamilton, L. S., Chang, D. L., Lee, M. B. & Chang, E. F. Semi-
automated anatomical labeling and inter-subject warping of high-
density intracranial recording electrodes in electrocorticography.
Front Neuroinform 11, 62 (2017).
65. Rolls, E. T., Huang, C.-C., Lin, C.-P., Feng, J. & Joliot, M. Automated
anatomical labelling atlas 3. Neuroimage 206, 116189 (2020).
66. Arora, P., Mishra, A. & Malhi, A. Machine learning Ensemble for the
Parkinson’s disease using protein sequences. Multimed. Tools Appl
81, 32215–32242 (2022).
Acknowledgements
We would like to present our acknowledgments to our patients for
participating in this project. JGZ is supported by the National Nature
Science Foundation of China (81830033) and the YangFan Project from
the Beijing Hosptial Management Center (ZLRK202313). ACY is
supported by the National Nature Science Foundation of China
(81870888).
Author contributions
Conception and design: ZXY, SL, LML, JGZ; Acquisition of data: ZXY, TSY,
GYZ, RYM, YCX, QA, YFG, GFQ, HTX, NZ, CXW, FGM, ACY; Analysis and
interpretation of data: ZXY, HLY, CS, MFA,TM, YJ, WJN, PS, SL, LML, JGZ;
First draft of manuscript: ZXY; Revision of manuscript: ZXY, HLY, CS, TM,
WJN, PS, SL, LML, JGZ.
Competing interests
WJN received honoraria for talks unrelated to this manuscript from
Medtronic which is a manufacturer of deep brain stimulation devices. The
remaining authors declare no competing interests.
Additional information
Supplementary information The online version contains
supplementary material available at
https://doi.org/10.1038/s41746-024-01115-7.
Correspondence and requests for materials should be addressed to
Zixiao Yin, Simon Little or Jianguo Zhang.
Reprints and permissions information is available at
http://www.nature.com/reprints
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons licence and your intended use is not permitted
by statutory regulation or exceeds the permitted use, you will need to
obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/.
© The Author(s) 2024
https://doi.org/10.1038/s41746-024-01115-7 Article
npj Digital Medicine | (2024) 7:122 13
