Created
June 24, 2023 12:56
-
-
Save DimitriPapadopoulos/599ccc16b294d22e833e566176ed3caf to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/python3 | |
from pathlib import Path | |
import numpy as np | |
from scipy.misc import electrocardiogram | |
from enum import IntEnum | |
from pyedflib import highlevel | |
class SleepStage(IntEnum): | |
""" | |
Mapping of AASM sleep stages to integers. | |
To facilitate hypnogram plotting, values start with zero and increase with wakefulness. | |
""" | |
UNDEFINED = 0 | |
N3 = 1 | |
N2 = 2 | |
N1 = 3 | |
REM = 4 | |
WAKE = 5 | |
def _dummy_nsrr_edf(filename: str, hours: float, ecg_channel: str): | |
ECG_FS = 360 | |
ecg_5_min = electrocardiogram() | |
seconds = int(hours * 60 * 60) | |
ecg = np.tile(ecg_5_min, int(np.ceil(seconds / 300)))[np.newaxis, : seconds * ECG_FS] | |
signal_headers = highlevel.make_signal_headers([ecg_channel], sample_frequency=ECG_FS) | |
highlevel.write_edf(filename, ecg, signal_headers) | |
def _dummy_nsrr_xml(filename: str, hours: float, random_state: int): | |
EPOCH_LENGTH = 30 | |
STAGES = [ | |
"Wake|0", | |
"Stage 1 sleep|1", | |
"Stage 2 sleep|2", | |
"Stage 3 sleep|3", | |
"Stage 4 sleep|4", | |
"REM sleep|5", | |
"Unscored|9", | |
] | |
rng = np.random.default_rng(random_state) | |
with open(filename, "w") as xml_file: | |
xml_file.write( | |
'<?xml version="1.0" encoding="UTF-8" standalone="no"?>\n' | |
"<PSGAnnotation>\n" | |
f"<EpochLength>{EPOCH_LENGTH}</EpochLength>\n" | |
"<ScoredEvents>\n" | |
"<ScoredEvent>\n" | |
"<EventType/>\n" | |
"<EventConcept>Recording Start Time</EventConcept>\n" | |
"<ClockTime>01.01.85 20.29.59</ClockTime>\n" | |
"</ScoredEvent>\n", | |
) | |
record_duration = hours * 60 * 60 | |
start = 0 | |
while True: | |
if start > record_duration: | |
break | |
epoch_duration = rng.choice(np.arange(4, 21)) * EPOCH_LENGTH | |
stage = rng.choice(STAGES) | |
xml_file.write( | |
"<ScoredEvent>\n" | |
"<EventType>Stages|Stages</EventType>\n" | |
f"<EventConcept>{stage}</EventConcept>\n" | |
f"<Start>{start:.1f}</Start>\n" | |
f"<Duration>{epoch_duration:.1f}</Duration>\n" | |
"</ScoredEvent>\n", | |
) | |
start += epoch_duration | |
xml_file.write( | |
"</ScoredEvents>\n" "</PSGAnnotation>\n", | |
) | |
def _create_dummy_shhs(data_dir: str, durations: list[float], random_state: int = 42): | |
DB_SLUG = "shhs" | |
ANNOTATION_DIRNAME = "polysomnography/annotations-events-nsrr" | |
EDF_DIRNAME = "polysomnography/edfs" | |
CSV_DIRNAME = "datasets" | |
db_dir = Path(data_dir).expanduser() / DB_SLUG | |
annotations_dir = db_dir / ANNOTATION_DIRNAME | |
edf_dir = db_dir / EDF_DIRNAME | |
csv_dir = db_dir / CSV_DIRNAME | |
for directory in (annotations_dir, edf_dir): | |
for visit in ("shhs1", "shhs2"): | |
(directory / visit).mkdir(parents=True, exist_ok=True) | |
csv_dir.mkdir(parents=True, exist_ok=True) | |
for i, hours in enumerate(durations): | |
for visit in ("shhs1", "shhs2"): | |
record_id = f"{visit}/{visit}-20{i:04}" | |
_dummy_nsrr_edf(f"{edf_dir}/{record_id}.edf", hours, ecg_channel="ECG") | |
_dummy_nsrr_xml(f"{annotations_dir}/{record_id}-nsrr.xml", hours, random_state) | |
with open(csv_dir / "shhs1-dataset-0.0.0.csv", "w") as csv: | |
csv.write("nsrrid,age_s1,gender,weight\n") | |
for i in range(len(durations)): | |
csv.write(f"2{i:05},55,1,77\n") | |
with open(csv_dir / "shhs2-dataset-0.0.0.csv", "w") as csv: | |
csv.write("nsrrid,age_s2,gender,weight\n") | |
for i in range(len(durations)): | |
csv.write(f"2{i:05},61,2,\n") | |
def main(): | |
durations = [0.1, 0.2] # hours | |
valid_stages = {int(s) for s in SleepStage} | |
_create_dummy_shhs(data_dir="/tmp", durations=durations) | |
if __name__ == '__main__': | |
main() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment