Error related-negativity (ERN) classification

Name: error-related negativity

Category: brain-computer interfacing

Dataset: Kappenman2021ErpErn

Objective: Binary classification

Split: Leave-subjects-out

Usage

neuralbench eeg ern

Show config.yaml

# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
#
# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.

data:
  study:
    source:
      name: Kappenman2021ErpErn
    filter_valid_events:
      name: QueryEvents
      query: "(type == 'Eeg') | (code.isin([1, 2]))"
    split:
      name: SklearnSplit
      split_by: subject
      valid_split_ratio: 0.2
      test_split_ratio: 0.2
      valid_random_state: 33
      test_random_state: 33
  neuro.baseline: [0.0, 0.2]
  target:
    =replace=: true
    name: LabelEncoder
    event_types: Stimulus
    event_field: description
    return_one_hot: true
    aggregation: trigger
  trigger_event_type: Stimulus
  start: 0.6
  duration: 1.0
  summary_columns: [description]
compute_class_weights: true
brain_model_output_size: &brain_model_output_size 2
trainer_config.monitor: val/bal_acc
trainer_config.mode: max
loss:
  name: CrossEntropyLoss
  kwargs:
    label_smoothing: 0.1
metrics: !!python/object/apply:neuralbench.defaults.metrics.get_classification_metric_configs
  - *brain_model_output_size

Description

The error related-negativity (ERN) classification task involves identifying error-related potentials in EEG recordings. ERNs are brain responses elicited when a person makes a mistake, often used in BCI applications. In this task, we use the Kappenman2021ErpErn dataset [Kappenman2021], part of the ERP CORE (Compendium of Open Resources and Experiments), which contains EEG data from 40 subjects performing a flanker task that elicits ERN responses on error trials.

Additional Datasets

The following additional datasets from MOABB can also be used with this task:

• Chavarriaga2010 (BNCI2015_013) – 6 subjects, observation-based ERN from a moving cursor paradigm

• Kueper2024 (IntEr-HRI) – 8 subjects, error-related potentials from orthosis-induced movement errors

To run with an alternate dataset:

neuralbench eeg ern --datasets kueper2024

References

[Kappenman2021]

Kappenman, E. S., et al. “ERP CORE: An open resource for human event-related potential research.” NeuroImage 225 (2021): 117465.