Module

The module is called ecg_gudb_database:

from ecg_gudb_database import GUDb

The constructor loads the ECG data of one subject/experiment from github:

ecg_class = GUDb(subject_number, experiment)

where subject_number is from 0..24 and experiment is ‘sitting’, ‘maths’, ‘walking’, ‘hand_bike’ or ‘jogging’. The array ecg_class.experiments is an array of all experiments so that one can loop through the different experiments.

Optionally, in case you decide later to download the whole dataset from http://researchdata.gla.ac.uk/716/ then specify the absolute path to the dataset with the optional parameter url without the “file:” specifier:

ecg_class = GUDb(subject_number, experiment, url = "/home/bp1/dataset_dataset_716/experiment_data/")

Retrieve the ECG data

The data is available as numpy arrays. The sampling rate is 250Hz for all experiments (ecg_class.fs). We have recorded Einthoven and from a chest strap.

Einthoven:

ecg_class.einthoven_I, ecg_class.einthoven_I_filt
ecg_class.einthoven_II, ecg_class.einthoven_II_filt
ecg_class.einthoven_III, ecg_class.einthoven_III_filt

Chest strap:

ecg_class.cs_V2_V1, ecg_class.cs_V2_V1_filt

where the filtered versions have 50Hz mains and DC removed.

R peak annotations

The two boolean variables ecg_class.anno_cs_exists and ecg_class.anno_cables_exists tell the user if annotations exist. If yes they can be obtained:

if ecg_class.anno_cs_exists:
    chest_strap_anno = ecg_class.anno_cs
else:
    print('No chest strap annotations')
if ecg_class.anno_cables_exists:
    cables_anno = ecg_class.anno_cables
else:
    print("No cables annotations")

Accelerometer data

The accelerometer was worn on a standard belt around the subject’s waist:

ecg_class.acc_x
ecg_class.acc_y
ecg_class.acc_z