How to change the EIT processing settings#

session.preprocess_eit() filters the EIT data, finds the breaths and calculates the results per breath. Give the settings as keyword arguments. The settings go to EITProcessingAdapter.preprocess.

Change the filter#

To select the filter, set filter_mode:

`filter_mode`	What it does
`"mdn"` (default)	Removes the heart signal with a multiple digital notch filter. The filter needs the respiratory rate and the heart rate, so M3Resp always calculates them for this mode.
`"lowpass"`	A Butterworth low-pass filter with the cutoff `lowpass_hz` (default 1.0 Hz).
`"bandpass"`	A Butterworth band-pass filter from `highpass_hz` (default 0.05 Hz) to `lowpass_hz` (default 1.0 Hz).
`"none"`	No filter.

filter_order sets the order of the low-pass and band-pass filters (default 4).

session.preprocess_eit(filter_mode="bandpass", highpass_hz=0.05, lowpass_hz=1.0)

Change the rate detection#

M3Resp finds the respiratory rate and the heart rate in the raw EIT data. If the patient is a neonate, set subject_type="neonate" (default "adult"). To change the length of the analysis window, set welch_window_seconds (default 30.0 s).

Change the breath detection#

M3Resp finds the breaths during preprocess_eit. To change the shortest breath that M3Resp accepts, set breath_min_duration_seconds (default 2/3 s).

detect_eit_breaths() then stores these breaths as BreathEvent objects in session.events["eit_breaths"].

Calculate fewer results#

Set a result to False to skip it:

Setting	Result
`compute_rates`	Respiratory rate and heart rate, in Hz
`compute_breath_intervals`	The start and end of each breath
`compute_continuous_tiv`	The tidal impedance variation (TIV) of each breath
`compute_eeli`	The end-expiratory lung impedance (EELI) of each breath
`compute_pixel_tiv`	A TIV map per breath, with one value per pixel

TIV, EELI and pixel TIV need the breaths. If you set compute_breath_intervals=False, also set the three TIV and EELI settings to False.

Keep more than one result#

To compare two settings, store each result under a name with variant:

session.preprocess_eit(variant="lowpass", filter_mode="lowpass")
session.detect_eit_breaths(variant="lowpass")

M3Resp stores the breaths of this result in session.events["eit_breaths:lowpass"].

Use the same settings with the preset#

Put the settings under "preprocess" in the preset configuration:

session.run_preset("eit", config={"preprocess": {"filter_mode": "lowpass"}})