psgscoring 0.2.9

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.

Validation (v0.2.8)

External validation on the PSG-IPA dataset (PhysioNet, 5 recordings, 47 independent scorer sessions):

PSG	Scorers	Median AHI	psgscoring AHI	ΔAHI	Severity
SN1	11	6.0	8.1	+2.1	Mild ✓
SN2	4	4.4	9.3	+4.9	Mild / Normal
SN3	12	54.0	53.8	−0.2	Severe ✓
SN4	12	3.5	4.3	+0.8	Normal ✓
SN5	12	9.9	11.4	+1.5	Mild ✓
Mean				1.9	4/5 concordant

Details: Online Supplement (Wiki)

Features

• AASM 2.6 respiratory scoring: apnea/hypopnea detection with dual-sensor support
• 12 systematic bias corrections: 6 over-counting + 6 under-counting (details)
• Breath-amplitude stability filter (v0.2.8): rejects false-positive hypopneas during normal breathing
• AHI confidence interval: strict/standard/sensitive profiles with robustness grade (A/B/C)
• Configurable scoring profiles: strict (research), standard (AASM 2.6), sensitive (UARS)
• ECG-derived effort classification: spectral + TECG for central apnea detection
• PLM, SpO₂, RERA/RDI, signal quality assessment

Installation

pip install psgscoring

Quick start

from psgscoring import run_pneumo_analysis
import mne

raw = mne.io.read_raw_edf("recording.edf", preload=True)
hypno = ["W", "N1", "N2", "N3", "R", ...]  # 30-s epochs

results = run_pneumo_analysis(raw, hypno, scoring_profile="standard")

# AHI
print(f"AHI: {results['respiratory']['summary']['ahi_total']}")

# AHI confidence interval (v0.2.8)
iv = results["ahi_interval"]
print(f"Interval: [{iv['interval'][0]} – {iv['interval'][1]}] Grade: {iv['robustness_grade']}")

What's new in v0.2.8

• Removed flow smoothing from standard profile (root cause: +54 false hypopneas)
• Stability filter: rejects hypopneas during stable breathing (CV < 0.45)
• AHI confidence interval: 3-profile analysis with robustness A/B/C
• PSG-IPA validation: mean |ΔAHI| = 1.9/h, severity concordance 4/5

Documentation

📖 Online Supplement (Wiki) — signal processing, corrections, validation
📖 Technical Handbook (PDF) — 26-page guide

Live platform

slaapkliniek.be — upload EDF, receive complete PSG report. No installation required.

Citation

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

This library builds on YASA:

Vallat R, Walker MP. An open-source, high-performance tool for automated sleep staging. eLife. 2021;10:e70092.

License

BSD-3-Clause. See LICENSE.

Disclaimer: Research use only. Not CE-marked or FDA-cleared. See DISCLAIMER.md.