Skip to main content

An interoperable and harmonized format for polysomnography recordings.

Project description

DOI PyPI version

sleeplab-format

Sleeplab format (SLF) is a both machine and human-readable format for storing and processing polysomnography data. SLF provides reader and writer with validation of data types and structures. The goal is to make it easier to apply analytics and machine learning pipelines to multiple datasets from different sources.

Documentation

See full documentation and the paper for details.

Related tools

sleeplab-converters for converting other formats exported from PSG software to sleeplab format.

sleeplab-tf-dataset for reading data in sleeplab format as a tensorflow Dataset.

Installation

pip install sleeplab-format

Usage

import sleeplab_format as slf
import pandas as pd
from pathlib import Path

# Read a toy dataset
DS_DIR = Path('tests/datasets/dataset1')
ds = slf.reader.read_dataset(DS_DIR)

# Get the list of subjects from series1
subjects = ds.series['series1'].subjects.values()

# Flatten the nested annotations and cast Pydantic models to dicts
all_events_dict = [dict(a)
    for s in subjects
    for a_model in s.annotations.values()
    for a in a_model.annotations
]

# Create a pandas DataFrame for analyses
event_df = pd.DataFrame(all_events_dict)

# Calculate the mean duration of hypopneas
hypopneas = event_df[event_df['name'] == 'HYPOPNEA']
mean_duration = sum(hypopneas['duration']) / len(hypopneas)

# Modify the dataset
additional_info = {'neck_size': 40.0}
ds.series['series1'].subjects['10001'].metadata.additional_info = additional_info
ds.name = 'dataset2'

# Write the modified dataset
MODIFIED_DS_DIR = Path('/tmp/datasets')
slf.writer.write_dataset(ds, MODIFIED_DS_DIR)

See the automatic sleep staging example for a full end-to-end example.

Contributing

See contributing.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

sleeplab_format-0.4.1.tar.gz (108.6 kB view hashes)

Uploaded Source

Built Distribution

sleeplab_format-0.4.1-py3-none-any.whl (16.3 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page