PANDA pediatric arousal neural detection architecture
12 affiliations
- 1Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA USA
- 2Harvard John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA USA
- 3Harvard Medical School, Boston, MA USA
- 4Emory University School of Medicine, Atlanta, GA USA
- 5Department of Epidemiology and Biostatistics, University of California, San Francisco, CA USA
- 6California Pacific Medical Center Research, San Francisco, CA USA
- 7Stanford University, Palo Alto, CA USA
- 8Kaiser Permanente, San Bernardino County Sleep Disorders Center, San Bernardino, CA USA
- 9Department of Neurology and Emory Sleep Center, Emory University School of Medicine, Atlanta, GA USA
- 10Georgia Institute of Technology, Atlanta, GA USA
- 11Pediatric Sleep Center, Boston Children’s Hospital, Boston, MA USA
- 12Department of Medicine, Division of Pulmonary Critical Care & Sleep Medicine, Beth Israel Deaconess Medical Center, Boston, MA USA
Abstract
We developed and validated PANDA, a deep-learning model for pediatric arousal detection in polysomnography. PANDA uses a U-Net-style encoder-decoder architecture trained on 10-channel PSG signals with 17.5 min contextual windows and 2 Hz arousal outputs. The model was trained on 9150 PSGs from 7604 subjects, validated on 2455 PSGs from 2000 subjects, and evaluated on a held-out Boston Children’s Hospital test set of 3804 recordings from 3000 subjects. To reduce reference-label noise, we constructed a rigorously adjudicated 200-PSG platinum set and quantified inter-rater reliability before and after adjudication. Performance was assessed using sample-wise agreement, event-wise arousal-index agreement, and external validation on CHAT and PATS. On routine labels, PANDA achieved mean subject-wise Cohen’s κ = 0.45 compared with κ = 0.26 for CAISR. Agreement increased to κ = 0.87 on platinum labels, consistent with reduced scoring variability after expert adjudication. PANDA also showed closer event-wise agreement with reference arousal index and lower bias than CAISR across internal and external cohorts. External validation yielded κ = 0.46 on CHAT and κ = 0.38 on PATS, improving to κ = 0.57 and κ = 0.50 after cohort-specific calibration. PANDA enables scalable pediatric arousal detection with improved consistency and reduced manual scoring time.
Related
- Beyond arousal: pupil fluctuations, neural activity, and behavior
- A Physics-Informed Neural Network Surrogate for Patient-Specific Hepatic Arterial Hemodynamics in Yttrium-90 Radioembolization: Network Architecture, Boundary-Condition Enforcement, and Data Efficiency
- Spatial correspondence between spontaneous brain activity alterations and normative molecular and neurochemical architecture in pediatric epilepsy.
- Universal rhythmic architecture uncovers two modes of neural dynamics
- Mnemonic Demand Reconfigures the Neural Architecture of Memory Encoding
BibTeX
@article{gupta2026panda,
title = {{PANDA pediatric arousal neural detection architecture}},
author = {Arnav Gupta and Ayush Tripathi and Wolfgang Ganglberger and Samuel Waters and Haoqi Sun and Samaneh Nasiri and Katie L. Stone and Emmanuel Mignot and Dennis Hwang and Matthew A. Reyna and Lynn Marie Trotti and Gari D. Clifford and Kiran Maski and Umakanth Katwa and Robert J. Thomas and M. Brandon Westover},
journal = {NPJ Digital Medicine},
year = {2026},
month = jul,
doi = {10.1038/s41746-026-02967-x},
url = {https://doi.org/10.1038/s41746-026-02967-x},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - PANDA pediatric arousal neural detection architecture
AU - Arnav Gupta
AU - Ayush Tripathi
AU - Wolfgang Ganglberger
AU - Samuel Waters
AU - Haoqi Sun
AU - Samaneh Nasiri
AU - Katie L. Stone
AU - Emmanuel Mignot
AU - Dennis Hwang
AU - Matthew A. Reyna
AU - Lynn Marie Trotti
AU - Gari D. Clifford
AU - Kiran Maski
AU - Umakanth Katwa
AU - Robert J. Thomas
AU - M. Brandon Westover
PY - 2026
DA - 2026/07/07
JO - NPJ Digital Medicine
DO - 10.1038/s41746-026-02967-x
UR - https://doi.org/10.1038/s41746-026-02967-x
AB - We developed and validated PANDA, a deep-learning model for pediatric arousal detection in polysomnography. PANDA uses a U-Net-style encoder-decoder architecture trained on 10-channel PSG signals with 17.5 min contextual windows and 2 Hz arousal outputs. The model was trained on 9150 PSGs from 7604 subjects, validated on 2455 PSGs from 2000 subjects, and evaluated on a held-out Boston Children’s Hospital test set of 3804 recordings from 3000 subjects. To reduce reference-label noise, we constructed a rigorously adjudicated 200-PSG platinum set and quantified inter-rater reliability before and after adjudication. Performance was assessed using sample-wise agreement, event-wise arousal-index agreement, and external validation on CHAT and PATS. On routine labels, PANDA achieved mean subject-wise Cohen’s κ = 0.45 compared with κ = 0.26 for CAISR. Agreement increased to κ = 0.87 on platinum labels, consistent with reduced scoring variability after expert adjudication. PANDA also showed closer event-wise agreement with reference arousal index and lower bias than CAISR across internal and external cohorts. External validation yielded κ = 0.46 on CHAT and κ = 0.38 on PATS, improving to κ = 0.57 and κ = 0.50 after cohort-specific calibration. PANDA enables scalable pediatric arousal detection with improved consistency and reduced manual scoring time.
SN - 2398-6352
ER -