Parkinson’s disease diagnosis classification

Name: parkinsons diagnosis

Category: clinical

Dataset: Singh2021

Objective: Multiclass classification

Split: Leave-subjects-out

Usage

neuralbench eeg parkinsons_diagnosis

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: Singh2021Timing
    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: diagnosis
  neuro:
    picks: [Fp1, Fz, F3, F7, FC5, FC1, C3, T7, TP9, CP5, CP1, P3, P7, O1, Oz, O2, P4, P8, TP10, CP6, CP2, Cz, C4, T8, FT10, FC6, FC2, F4, F8, Fp2, AF7, AF3, AFz, F1, F5, FT7, FC3, C1, C5, TP7, CP3, P1, P5, PO7, POz, PO8, P6, P2, CPz, CP4, TP8, C6, C2, FC4, FT8, F6, AF8, AF4, F2, FCz]
    # Not available for every timeline: Iz, I1, I2
  target:
    =replace=: true
    name: LabelEncoder
    event_types: Eeg
    event_field: diagnosis
    return_one_hot: true
    aggregation: trigger
  trigger_event_type: Eeg
  start: 0.0
  duration: 5.0
  stride: 5.0
  summary_columns: [diagnosis]
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
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

This task consists in predicting whether a subject has received a Parkinson’s disease (PD) diagnosis or is a healthy control (HC) based on EEG recordings during an interval timing task, where participants had to press a key after estimating a time interval of either 3 or 7 seconds.

Dataset Notes

• We use the re-packaged version of Singh2021 by [Kastrati2025], available on HuggingFace. Sessions with insufficient data or poor signal quality, or with fewer than 20 valid key presses per interval condition were excluded from analyses [Singh2021].

References

[Singh2021]

Singh, Arun, et al. “Timing variability and midfrontal~ 4 Hz rhythms correlate with cognition in Parkinson’s disease.” npj Parkinson’s Disease 7.1 (2021): 14.

[Kastrati2025]

Kastrati, Ard, et al. “EEG-Bench: A Benchmark for EEG Foundation Models in Clinical Applications.” arXiv preprint arXiv:2512.08959 (2025).