Parameters#

A parameter is a number that M3Resp calculates from the data, for example the respiratory rate, the EMG amplitude or the delay between the EIT and the EMG breath. M3Resp stores each parameter as a ParameterResult object in session.parameter_results. The value can be one number or an array.

Look at a parameter#

rate = session.parameter_results.for_name("respiratory_rate")[0]
print(rate.value)     # 0.2313914079219103
print(rate.unit)      # Hz
print(rate.modality)  # eit
print(rate.method)    # eitprocessing.RateDetection

To find parameters, use for_name, for_modality or for_category:

session.parameter_results.for_modality("emg")

What a parameter holds#

Field	Contents
`name`	The name of the parameter, for example `"respiratory_rate"`
`value`	One number, or an array
`modality`	The device that the parameter comes from
`category`	The physical quantity that the parameter is calculated from (see Signals)
`unit`	The unit, for example `"Hz"` or `"s"`
`metric_type`	The standard name of the parameter. M3Resp sets it from `name` for a known parameter, and leaves it `None` for a name of your own.
`breath_id`, `breath_ids`, `event_id`, `start_time`, `end_time`	The breath, event, or time window the value refers to (see below)
`region`, `channel`	The lung region or the channel of the value
`method`	The calculation that made the value
`metadata`	Other details, as a dictionary

For example, a parameter with name="RR" and one with name="breath_rate" both get metric_type="respiratory_rate", so you can group them. A parameter with name="mean_amplitude" gets metric_type=None.

is_scalar is True when the value is one number. to_dict() converts the contents of the ParameterResult object to a dictionary that can be written into a JSON file, with arrays serialized to lists.

What a value belongs to#

The value of a parameter can belong to the whole recording or to a part of it. Set the fields that apply. You can combine them, for example one value per breath in a time window. The one exception is breath_id together with breath_ids: a parameter with both gives a ValueError.

Fields	The value belongs to
None of the fields	The whole recording
`breath_id`	One breath
`breath_ids`	Several breaths, for example a mean over five breaths
`event_id`	One event, for example a blood gas sample
`start_time`	One time point, in seconds
`start_time` and `end_time`	One time window, in seconds. For several windows, make one parameter per window.

from m3resp import ParameterResult

mean_amplitude = ParameterResult(
    name="mean_amplitude",
    value=3.8,
    modality="emg",
    start_time=30.0,
    end_time=60.0,
)

Values with one value per breath, such as the TIV and EELI of the EIT, are stored as values per breath in session.interval_data. Each value then stays next to its breath.

Where parameters come from#

Call	Parameters
`preprocess_eit`	The respiratory rate and the heart rate of the EIT, in Hz
`postprocess_emg` (or the `"emg"` preset)	The EMG parameters, for example the amplitude, the time product and the respiratory rate
`compute_breath_timing_parameters`	The delay and the difference in duration between modalities, and the fraction of breaths that both modalities found (see Synchronization and breath timing)
Workflow steps	Many steps add parameters, for example the ECG gate mask of `emg.ecg_gating`

The timing parameters between modalities use the start, extremum and end times of the breaths. Each modality marks these times at its own moment of the breath, so use the delays to check the detection and the synchronization.

session.export_summary writes the parameters to parameter_results.csv (see How to export the results of a session).

Continue with Quality flags.