Seizure detection

Name: seizure
Category: clinical
Dataset: Dan2023 (CHB-MIT)
Objective: Multiclass classification
Split: Leave-subjects-out

Usage

neuralbench eeg seizure
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: Dan2023Bids
    split:
      name: SklearnSplit
      split_by: subject
      valid_split_ratio: 0.2
      test_split_ratio: 0.2
      valid_random_state: 33
      test_random_state: 33
      stratify_by: state
  channel_positions:
    layout_or_montage_name: standard_1005
  target:
    =replace=: true
    name: LabelEncoder
    event_types: Seizure
    event_field: state
    return_one_hot: true
    aggregation: sum
    allow_missing: true
    treat_missing_as_separate_class: true
  trigger_event_type: Eeg
  start: 0.0
  duration: 5.0
  stride: 5.0
  stride_drop_incomplete: true
  use_weighted_sampler: true
  summary_columns: [state]
compute_class_weights: false
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
test_full_metrics:
  - log_name: acc
    name: Accuracy
    kwargs:
      task: multiclass
      num_classes: *brain_model_output_size
  - log_name: bal_acc
    name: Accuracy
    kwargs:
      task: multiclass
      num_classes: *brain_model_output_size
      average: macro
  - log_name: auroc
    name: AUROC
    kwargs:
      task: multiclass
      num_classes: *brain_model_output_size

Description

The seizure detection task involves identifying epileptic seizures from EEG recordings [Shoeb2009]. The dataset used for this task is the CHB-MIT dataset, which contains EEG data from pediatric subjects with intractable seizures [Guttag2010]. Here, we use the BIDS-compatible version of the dataset [Dan2023]. We combine all seizure types into a single positive class for this binary classification task.

Dataset Notes

  • Subject 21 in the original study [Guttag2010] was actually the second session of subject 1.

References

[Dan2023]

Dan, J.and A. Shoeb. BIDS CHB-MIT Scalp EEG Database. v1.0.0, EPFL, 5 Dec. 2023, https://doi.org/10.5281/zenodo.10259996.

[Guttag2010] (1,2)

Guttag, J. (2010). CHB-MIT Scalp EEG Database (version 1.0.0). PhysioNet. RRID:SCR_007345. https://doi.org/10.13026/C2K01R

[Shoeb2009]

Shoeb, Ali Hossam. Application of machine learning to epileptic seizure onset detection and treatment. Diss. Massachusetts Institute of Technology, 2009.