psgscoring 0.2.6

Open-source Python library for AASM 2.6-compliant automated polysomnography scoring

psgscoring

Open-source Python library for AASM 2.6-compliant automated polysomnography scoring.

psgscoring extracts the core respiratory scoring algorithms from YASAFlaskified into a standalone, pip-installable library for the research community.

Documentation

📖 Technical Handbook (PDF) — 26-page guide covering:

• Clinical context and AASM 2.6 scoring rules
• Signal processing pipeline (linearisation, filtering, Hilbert envelope, dynamic baseline)
• 7-rule apnea type classification with Hilbert phase-angle analysis
• All 12 over-counting and under-counting corrections with pseudocode
• ECG-derived effort classification (spectral + TECG)
• Arousal detection, RERA/RDI, snoring, heart rate, body position, Cheyne-Stokes
• PLM detection, signal quality assessment, sleep cycle analysis
• Complete pipeline walkthrough with code examples
• Data structures, testing strategy, design decisions
• Glossary and exercises

Written for BSc-level computer science students. LaTeX source included in docs/handbook.tex.

Features

• AASM 2.6 respiratory scoring: apnea/hypopnea detection with dual-sensor (thermistor + nasal pressure) support
• 7-rule apnea type classification: obstructive / central / mixed with Hilbert phase-angle analysis
• 6 over-counting corrections: post-apnea baseline inflation, SpO₂ cross-contamination, Cheyne-Stokes flagging, borderline classification, artefact-flank exclusion, local baseline validation
• 6 under-counting corrections: peak-based hypopnea detection, SpO₂ de-blocking, extended nadir window, flow smoothing, position auto-mapping, configurable scoring profiles
• ECG-derived effort classification: adaptive cardiac band + TECG (Berry 2019) for central apnea detection
• PLM scoring per AASM 2.6 + WASM criteria
• SpO₂ analysis: ODI 3%/4%, baseline, T90, low-baseline warning
• Signal quality assessment: flat-line, clipping, disconnect, montage plausibility
• RERA/RDI computation: amplitude-reduction + flattening-based RERA detection

Installation

pip install psgscoring

With optional dependencies (YASA staging + LightGBM):

pip install psgscoring[full]

Quick start

import numpy as np
from psgscoring import (
    bandpass_flow, linearise_nasal_pressure,
    compute_dynamic_baseline, classify_apnea_type,
    detect_breaths, compute_flattening_index,
)

# Load your flow signal (e.g., from MNE)
# flow = raw.get_data(picks='Flow')[0]

# Linearise nasal pressure (Bernoulli correction)
flow_lin = linearise_nasal_pressure(flow)

# Bandpass filter (0.05–3 Hz)
flow_filt = bandpass_flow(flow_lin, sf=256)

# Dynamic baseline (5-min sliding 95th percentile)
baseline = compute_dynamic_baseline(np.abs(flow_filt), sf=256)

# Detect breaths
breaths = detect_breaths(flow_filt, sf=256)

Submodules

Module	Responsibility
constants	AASM thresholds, band limits
utils	Sleep mask, channel detection
signal	Linearisation, baseline, MMSD
breath	Breath segmentation, flattening index
classify	Apnea type (7-rule + Hilbert)
spo2	SpO₂ coupling, ODI
plm	PLM detection (AASM 2.6 + WASM)
ecg_effort	ECG-derived effort (TECG + spectral)
ancillary	HR, snore, position, CSR
respiratory	Apnea/hypopnea + 12 corrections
pipeline	Master run_pneumo_analysis()
signal_quality	Per-channel quality grading

Configurable scoring profiles

Profile	Hypopnea threshold	Nadir window	Smoothing	Peak detection
Strict	≥30%	30 s	—	No
Standard	≥30%	45 s	3 s	Yes
Sensitive	≥25%	45 s	5 s	Yes

References

Standards and Scoring Manuals

1. Berry RB, Brooks R, Gamaldo C, et al. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Version 2.6. AASM, 2020.
2. Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events. 1st ed. AASM, 2007.
3. Zucconi M, Ferri R, Allen R, et al. WASM standards for recording and scoring periodic leg movements in sleep and wakefulness. Sleep Med. 2006;7(2):175–183. doi:10.1016/j.sleep.2006.01.001

Sleep Staging and YASA

4. Vallat R, Walker MP. An open-source, high-performance tool for automated sleep staging. eLife. 2021;10:e70092. doi:10.7554/eLife.70092
5. Perslev M, Darkner S, Kempfner L, et al. U-Sleep: resilient high-frequency sleep staging. npj Digital Medicine. 2021;4:72. doi:10.1038/s41746-021-00440-5
6. Ke G, Meng Q, Finley T, et al. LightGBM: a highly efficient gradient boosting decision tree. NeurIPS. 2017;30:3146–3154. NeurIPS 2017
7. Gramfort A, Luessi M, Larson E, et al. MNE software for processing MEG and EEG data. NeuroImage. 2014;86:446–460. doi:10.1016/j.neuroimage.2013.10.027
8. Feinberg I, Floyd TC. Systematic trends across the night in human sleep cycles. Psychophysiology. 1979;16(3):283–291. doi:10.1111/j.1469-8986.1979.tb02991.x

Respiratory Scoring and Signal Processing

9. Thurnheer R, Xie X, Bloch KE. Accuracy of nasal cannula pressure recordings for assessment of ventilation during sleep. Am J Respir Crit Care Med. 2001;164(10):1914–1919. doi:10.1164/ajrccm.164.10.2010113
10. Montserrat JM, Farré R, Ballester E, et al. Evaluation of nasal prongs for estimating nasal flow. Am J Respir Crit Care Med. 1997;155(1):211–215. doi:10.1164/ajrccm.155.1.9001310
11. Lee H, Park J, Kim H, et al. Detection of apneic events from single-channel nasal airflow using 2nd derivative method. Physiol Meas. 2008;29:N37–N45. doi:10.1088/0967-3334/29/5/N01
12. Hosselet J, Norman RG, Ayappa I, Rapoport DM. Detection of flow limitation with a nasal cannula/pressure transducer system. Am J Respir Crit Care Med. 1998;157(5):1461–1467. doi:10.1164/ajrccm.157.5.9708008
13. Uddin MB, Chow CM, Su SW. A novel algorithm for automatic diagnosis of sleep apnea from airflow and oximetry signals. Physiol Meas. 2021;42:015001. doi:10.1088/1361-6579/abd47a
14. Nakano H, Tanigawa T, Furukawa T, Nishima S. New rule-based algorithm for real-time detecting sleep apnea events. J Clin Sleep Med. 2016;12(10):1389–1396. doi:10.5664/jcsm.6200

Apnea Classification and Effort Assessment

15. Berry RB, Albertario CL, Guttman B, et al. Use of a transformed ECG signal to detect respiratory effort during apnea. J Clin Sleep Med. 2019;15(11):1653–1660. doi:10.5664/jcsm.7880
16. Berry RB, Budhiraja R, Guttman B, et al. Use of chest wall electromyography to detect respiratory effort during polysomnography. J Clin Sleep Med. 2016;12(9):1239–1244. doi:10.5664/jcsm.6122

Scoring Variability and Validation

17. Rosenberg RS, Van Hout S. The AASM inter-scorer reliability program: sleep stage scoring. J Clin Sleep Med. 2013;9(1):81–87. doi:10.5664/jcsm.2350
18. Rosenberg RS, Hirshkowitz M, Engleman HM, Penzel T. The AASM inter-scorer reliability program: respiratory events. J Clin Sleep Med. 2014;10(4):447–454. doi:10.5664/jcsm.3630
19. Ruehland WR, Rochford PD, O'Donoghue FJ, et al. The new AASM criteria for scoring hypopneas: impact on the apnea hypopnea index. Sleep. 2009;32(2):150–157. doi:10.1093/sleep/32.2.150
20. Malhotra A, Younes M, Kuna ST, et al. Performance of an automated polysomnography scoring system versus computer-assisted manual scoring. Sleep. 2013;36(4):573–582. doi:10.5665/sleep.2548

Novel Metrics

21. Parekh A, Kam K, Wickramaratne S, et al. Ventilatory burden as a measure of obstructive sleep apnea severity is predictive of cardiovascular and all-cause mortality. Am J Respir Crit Care Med. 2023;208(11):1216–1226. doi:10.1164/rccm.202301-0109OC

Epidemiology

22. Peppard PE, Young T, Barnet JH, et al. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013;177(9):1006–1014. doi:10.1093/aje/kws342
23. Benjafield AV, Ayas NT, Eastwood PR, et al. Estimation of the global prevalence and burden of obstructive sleep apnoea. Lancet Respir Med. 2019;7(8):687–698. doi:10.1016/S2213-2600(19)30198-5

Software

24. Vallat R. Pingouin: statistics in Python. J Open Source Software. 2018;3(31):1026. doi:10.21105/joss.01026

⚠️ Disclaimer

psgscoring is a research tool, not a certified medical device.

• This software is not CE-marked (MDR 2017/745), FDA-cleared, or approved by any regulatory authority for clinical diagnostic use.
• All scoring results must be reviewed and validated by a qualified sleep physician before any clinical decision.
• The automated AHI and event classifications carry inherent uncertainty. Confidence scores, correction counters, and scoring profiles are provided to help clinicians assess the robustness of each individual study — they do not replace expert judgment.
• The developers accept no liability for clinical decisions based on psgscoring output.
• Sleep staging depends on YASA's LightGBM classifier (~85% epoch agreement); staging errors propagate into all downstream indices.

For the full disclaimer, see DISCLAIMER.md.

Citation

If you use psgscoring in your research, please cite:

Rombaut B, Rombaut B, Rombaut C, Vallat R*. psgscoring: An Open-Source Python Library for AASM 2.6-Compliant Automated Polysomnography Scoring. 2026. https://github.com/bartromb/psgscoring

* Co-authorship subject to confirmation by R. Vallat

License

BSD-3-Clause. See LICENSE.