Segmentation¶

Bases: BaseClass

Class for segmentation (Epoching).

Extracts events from the specified stimulation channel and creates epochs around selected markers using MNE's find_events() and Epochs functions.

MNE-functions used: mne.find_events, mne.pick_events

Parameters:

Name	Type	Description	Default
`raw`	`Raw`	Input data.	required
`stim_channel`	`str \| None`	Name of the stimulation channel used to detect events (e.g., `'Status'`). If None, MNE falls back to config variables `'MNE_STIM_CHANNEL'`, `'STI 014'`, or the first `'stim'`-type channel found.	`None`
`start`	`float`	Start time of each epoch in seconds relative to the event.	required
`end`	`float`	End time of each epoch in seconds relative to the event	required
`baseline`	`tuple of (float, float) \| None`	Time interval for baseline correction passed to MNE Epochs. Use None to disable baseline correction.	`None`
`reference_markers`	`dict`	Marker selection configuration with the following optional keys: - "list" : list of int \| None Explicit marker IDs to include. - "range" : str \| None Inclusive marker range as `"min-max"` (e.g. `"10-69"`). If both keys are provided, their union is used. If None or empty, all detected events are used.	`None`
`eventsfile`	`str \| None`	Path to a CSV file defining event labels and their corresponding IDs or ID ranges separated with '-'. The CSV file should have two columns: "label" and "ID".	`None`
`**kwargs`		Additional keyword arguments passed to MNE's `find_events()` and `Epochs()` functions. Pass arguments for the different functions by using a dictionary with `find_events` and `Epochs` as keys.	`{}`

Attributes:

Name	Type	Description
`processed`	`Epochs`	Epoched output data.

Example settings for config file¶

Example settings in correct format for the config file.

Segmentation:
  _biopap_module: <class 'biopap.modules.segmentation.Segmentation'>
  stim_channel: 'Status'
  start: -0.2
  end: 1
  baseline: None
  reference_markers:
    list:
    - '100'
    - '101'
    - '120'
    range: 10-69
  eventsfile: /path/to/your/events_metadata.csv