{"dataset":{"id":"299","dataset_id":"nm000266","name":"Sosulski et al. 2019 — Online Optimization of Stimulation Speed in an Auditory Brain-Computer Interface under Time Constraints","description":"An auditory event-related potential dataset from 13 healthy subjects performing an oddball paradigm with two sinusoidal tones (target 1000 Hz, non-target 500 Hz) presented at variable stimulus onset asynchronies (60-600 ms). The dataset comprises 31-channel EEG recordings at 1000 Hz acquired with BrainProducts BrainAmp DC during an online closed-loop brain-computer interface experiment designed to optimize stimulation parameters using Bayesian optimization and random search strategies.","owner_user_id":19,"status":"active","github_repo":"nemarDatasets/nm000266","concept_doi":"10.82901/nemar.nm000266","latest_version_doi":"10.82901/nemar.nm000266.v1.0.4","created_at":"2026-03-26 18:40:56","updated_at":"2026-08-18 18:20:46","zenodo_concept_id":"20523718","is_sandbox":0,"visibility":"public","ezid_status":"public","enrichment_json":"{\n  \"version\": \"2.0\",\n  \"pipeline_stage\": \"validated\",\n  \"title\": \"Sosulski et al. 2019 — Online Optimization of Stimulation Speed in an Auditory Brain-Computer Interface under Time Constraints\",\n  \"description\": \"An auditory event-related potential dataset from 13 healthy subjects performing an oddball paradigm with two sinusoidal tones (target 1000 Hz, non-target 500 Hz) presented at variable stimulus onset asynchronies (60-600 ms). The dataset comprises 31-channel EEG recordings at 1000 Hz acquired with BrainProducts BrainAmp DC during an online closed-loop brain-computer interface experiment designed to optimize stimulation parameters using Bayesian optimization and random search strategies.\",\n  \"methods_description\": \"EEG data were recorded from 13 healthy subjects (age 20-26 years) using a 31-channel BrainProducts BrainAmp DC system at 1000 Hz sampling rate with passive Ag/AgCl electrodes in a standard 10-20 montage, referenced to nose. Subjects performed an auditory oddball task with target (1000 Hz) and non-target (500 Hz) tones presented via speaker at 65 cm distance. Each trial consisted of 90 stimuli (15 target, 75 non-target) in pseudo-random order. The experiment comprised an optimization phase (four strategies with 20-minute time windows) and a validation phase (20 trials per stimulus onset asynchrony). Classification used regularized linear discriminant analysis with mean amplitude features extracted from five time intervals (100-500 ms) across all channels.\",\n  \"license\": \"CC-BY-SA-4.0\",\n  \"dataset_type\": \"derivative\",\n  \"authors\": {\n    \"Jan Sosulski\": {\n      \"orcid\": \"0000-0002-8105-3395\"\n    },\n    \"David Hübner\": {},\n    \"Aaron Klein\": {},\n    \"Michael Tangermann\": {\n      \"orcid\": \"0000-0001-6729-0290\"\n    }\n  },\n  \"keywords\": [\n    {\n      \"term\": \"brain-computer interfaces\"\n    },\n    {\n      \"term\": \"EEG\"\n    },\n    {\n      \"term\": \"Event-Related Potentials, P300\",\n      \"subject_scheme\": \"MeSH\",\n      \"scheme_uri\": \"https://id.nlm.nih.gov/mesh/\",\n      \"value_uri\": \"http://id.nlm.nih.gov/mesh/D018913\"\n    },\n    {\n      \"term\": \"auditory oddball paradigm\"\n    },\n    {\n      \"term\": \"Bayesian optimization\"\n    },\n    {\n      \"term\": \"stimulus onset asynchrony\"\n    },\n    {\n      \"term\": \"closed-loop parameter optimization\"\n    }\n  ],\n  \"related_identifiers\": [\n    {\n      \"identifier\": \"10.6094/UNIFR/154576\",\n      \"identifier_type\": \"DOI\",\n      \"relation_type\": \"IsDerivedFrom\"\n    },\n    {\n      \"identifier\": \"10.48550/arXiv.2109.06011\",\n      \"identifier_type\": \"DOI\",\n      \"relation_type\": \"IsDerivedFrom\"\n    },\n    {\n      \"identifier\": \"https://github.com/nemarDatasets/nm000266\",\n      \"identifier_type\": \"URL\",\n      \"relation_type\": \"IsDescribedBy\"\n    },\n    {\n      \"identifier\": \"10.21105/joss.01896\",\n      \"identifier_type\": \"DOI\",\n      \"relation_type\": \"IsReferencedBy\"\n    },\n    {\n      \"identifier\": \"10.1038/s41597-019-0104-8\",\n      \"identifier_type\": \"DOI\",\n      \"relation_type\": \"IsReferencedBy\"\n    },\n    {\n      \"identifier\": \"https://nemar.org/dataset/nm000266\",\n      \"identifier_type\": \"URL\",\n      \"relation_type\": \"IsDescribedBy\"\n    }\n  ],\n  \"funding_references\": [\n    {\n      \"funder_name\": \"DFG\",\n      \"award_number\": \"EXC 1086\",\n      \"award_title\": \"Cluster of Excellence BrainLinks-BrainTools\"\n    },\n    {\n      \"funder_name\": \"DFG\",\n      \"award_number\": \"TA 1258/1-1\",\n      \"award_title\": \"DFG project SuitAble\"\n    },\n    {\n      \"funder_name\": \"DFG\",\n      \"award_number\": \"INST 39/963-1 FUGG\"\n    }\n  ],\n  \"resource_type_general\": \"Dataset\",\n  \"resource_type_specific\": \"EEG Dataset\",\n  \"modalities\": [\n    \"eeg\"\n  ],\n  \"sizes\": [\n    \"3.9 GB (2121 files)\"\n  ],\n  \"formats\": [\n    \".bdf\",\n    \".json\",\n    \".md\",\n    \".tsv\",\n    \".vhdr\",\n    \".yaml\",\n    \".yml\"\n  ],\n  \"source_hash\": \"ee980bd712bf22c12768d5a6d14259b36252de895baa62f9aea5181b2510ed93\"\n}","last_activity_at":"2026-08-16 14:03:36","source":null,"source_id":null,"subject_count":13,"modalities":"eeg","age_min":22.7,"age_max":22.7,"file_size":3948988656,"total_files":12731,"tasks":"p300","metadata_columns_error":null,"staleness_warn_stage":null,"staleness_admin_notified_at":null,"authors":"Jan Sosulski, David Hübner, Aaron Klein, Michael Tangermann","license":"CC-BY-SA-4.0","readme":"[![DOI](https://img.shields.io/badge/DOI-10.82901%2Fnemar.nm000266-blue)](https://doi.org/10.82901/nemar.nm000266)\n\nSosulski2019\n============\n\nP300 dataset from initial spot study.\n\nDataset Overview\n----------------\n  Code: Sosulski2019\n  Paradigm: p300\n  DOI: 10.6094/UNIFR/154576\n  Subjects: 13\n  Sessions per subject: 80\n  Events: Target=21, NonTarget=1\n  Trial interval: [-0.2, 1] s\n  File format: brainvision\n\nAcquisition\n-----------\n  Sampling rate: 1000.0 Hz\n  Number of channels: 31\n  Channel types: eeg=31, eog=1, misc=5\n  Channel names: C3, C4, CP1, CP2, CP5, CP6, Cz, EOGvu, F10, F3, F4, F7, F8, F9, FC1, FC2, FC5, FC6, Fp1, Fp2, Fz, O1, O2, P10, P3, P4, P7, P8, P9, Pz, T7, T8, x_EMGl, x_GSR, x_Optic, x_Pulse, x_Respi\n  Montage: standard_1020\n  Hardware: BrainProducts BrainAmp DC\n  Reference: nose\n  Sensor type: passive Ag/AgCl\n  Line frequency: 50.0 Hz\n  Auxiliary channels: EOG (1 ch, vertical)\n\nParticipants\n------------\n  Number of subjects: 13\n  Health status: healthy\n  Age: mean=22.7, std=1.64, min=20, max=26\n  Gender distribution: male=5, female=8\n  Species: human\n\nExperimental Protocol\n---------------------\n  Paradigm: p300\n  Number of classes: 2\n  Class labels: Target, NonTarget\n  Study design: Subjects focused attention on target tones (1000 Hz) and ignored non-target tones (500 Hz) presented via speaker at 65 cm distance. One trial consisted of 15 target and 75 non-target stimuli in pseudo-random order with at least two non-target tones between target tones. The experiment was split into optimization and validation parts.\n  Stimulus type: oddball\n  Stimulus modalities: auditory\n  Primary modality: auditory\n  Synchronicity: synchronous\n  Mode: online\n  Instructions: Focus on the target tones (1000 Hz) and ignore the non-target tones (500 Hz). Refrain from blinking and movement as much as possible.\n  Stimulus presentation: target_tone_hz=1000, non_target_tone_hz=500, tone_duration_ms=40, distance_cm=65\n\nHED Event Annotations\n---------------------\n  Schema: HED 8.4.0 | Browse: https://www.hedtags.org/hed-schema-browser\n\n  Target\n    ├─ Sensory-event\n    ├─ Experimental-stimulus\n    ├─ Visual-presentation\n    └─ Target\n\n  NonTarget\n    ├─ Sensory-event\n    ├─ Experimental-stimulus\n    ├─ Visual-presentation\n    └─ Non-target\n\nParadigm-Specific Parameters\n----------------------------\n  Detected paradigm: p300\n  Number of targets: 1\n\nData Structure\n--------------\n  Trials: Variable: optimization part used time-limited trials (20 minutes per strategy), validation part used 20 trials per SOA\n  Trials per class: target=13 per trial (after preprocessing, originally 15), non_target=65 per trial (after preprocessing, originally 75)\n  Trials context: Each trial consisted of 90 stimuli (15 target, 75 non-target). After preprocessing (removing first and last 6 epochs), 78 data points available per trial: 13 target and 65 non-target epochs.\n\nSignal Processing\n-----------------\n  Classifiers: rLDA, Shrinkage LDA\n  Feature extraction: Mean amplitude in time intervals\n  Frequency bands: analyzed=[1.5, 40.0] Hz\n\nCross-Validation\n----------------\n  Method: 13-fold\n  Folds: 13\n  Evaluation type: within_session\n\nPerformance (Original Study)\n----------------------------\n  Auc: 0.701\n  Mean Auc Ucb: 0.701\n  Mean Auc Rand: 0.704\n  Mean Auc P300 Ucb: 0.67\n  Mean Auc P300 Rand: 0.681\n  Mean Auc Fixed60: 0.517\n\nBCI Application\n---------------\n  Applications: communication\n  Online feedback: False\n\nTags\n----\n  Pathology: Healthy\n  Modality: Auditory\n  Type: Research\n\nDocumentation\n-------------\n  Description: Auditory oddball ERP dataset from 13 healthy subjects. Two sinusoidal tones (target 1000 Hz, non-target 500 Hz) presented at various stimulus onset asynchronies (SOAs, 60-600 ms). 31-channel EEG recorded at 1000 Hz with BrainProducts BrainAmp DC. Raw BrainVision format data.\n  DOI: 10.48550/arXiv.2109.06011\n  License: CC-BY-SA-4.0\n  Investigators: Jan Sosulski, David Hübner, Aaron Klein, Michael Tangermann\n  Senior author: Michael Tangermann\n  Contact: jan.sosulski@blbt.uni-freiburg.de; davhuebn@gmail.com; kleinaa@cs.uni-freiburg.de; michael.tangermann@donders.ru.nl\n  Institution: University of Freiburg\n  Country: DE\n  Repository: FreiDok\n  Data URL: https://freidok.uni-freiburg.de/data/154576\n  Publication year: 2021\n  Funding: Cluster of Excellence BrainLinks-BrainTools funded by the German Research Foundation (DFG) [grant number EXC 1086]; DFG project SuitAble [grant number TA 1258/1-1]; state of Baden-Württemberg, Germany, through bwHPC and the German Research Foundation (DFG) [grant number INST 39/963-1 FUGG]\n  Ethics approval: Approved by the ethics committee of the university medical center of Freiburg\n  Acknowledgements: Experiments were performed according to the Declaration of Helsinki.\n  Keywords: Bayesian optimization, individual experimental parameters, brain-computer interfaces, learning from small data, auditory event-related potentials, closed-loop parameter optimization\n\nAbstract\n--------\nThe decoding of brain signals recorded via, e.g., an electroencephalogram, using machine learning is key to brain-computer interfaces (BCIs). Stimulation parameters or other experimental settings of the BCI protocol typically are chosen according to the literature. The decoding performance directly depends on the choice of parameters, as they influence the elicited brain signals and optimal parameters are subject-dependent. Thus a fast and automated selection procedure for experimental parameters could greatly improve the usability of BCIs. We evaluate a standalone random search and a combined Bayesian optimization with random search into a closed-loop auditory event-related potential protocol. We aimed at finding the individually best stimulation speed—also known as stimulus onset asynchrony (SOA)—that maximizes the classification performance of a regularized linear discriminant analysis.\n\nMethodology\n-----------\nThe experiment was divided into two parts: (1) Optimization part: four strategies (AUC-ucb, AUC-rand, P300-ucb, P300-rand) each allocated 20 minutes to find optimal SOA. Strategies alternated to minimize non-stationarity effects. (2) Validation part: evaluated SOAs from each optimization strategy plus fixed 60ms SOA using 20 trials each (in blocks of 5 trials). Features were mean amplitudes in 5 time intervals ([100, 170], [171, 230], [231, 300], [301, 410], [411, 500] ms) across 31 channels (155 dimensions total). Classification used rLDA with automatic shrinkage regularization and 13-fold cross-validation on single trials.\n\nReferences\n----------\nSosulski, J., Tangermann, M.: Electroencephalogram signals recorded from 13 healthy subjects during an auditory oddball paradigm under different stimulus onset asynchrony conditions. Dataset. DOI: 10.6094/UNIFR/154576\n\nSosulski, J., Tangermann, M.: Spatial filters for auditory evoked potentials transfer between different experimental conditions. Graz BCI Conference. 2019.\n\nSosulski, J., Hübner, D., Klein, A., Tangermann, M.:  Online Optimization of Stimulation Speed in an Auditory Brain-Computer Interface under Time Constraints. arXiv preprint. 2021.\n\nNotes\n\n.. versionadded:: 0.4.5\nAppelhoff, S., Sanderson, M., Brooks, T., Vliet, M., Quentin, R., Holdgraf, C., Chaumon, M., Mikulan, E., Tavabi, K., Hochenberger, R., Welke, D., Brunner, C., Rockhill, A., Larson, E., Gramfort, A. and Jas, M. (2019). MNE-BIDS: Organizing electrophysiological data into the BIDS format and facilitating their analysis. Journal of Open Source Software 4: (1896). https://doi.org/10.21105/joss.01896\n\nPernet, C. R., Appelhoff, S., Gorgolewski, K. J., Flandin, G., Phillips, C., Delorme, A., Oostenveld, R. (2019). EEG-BIDS, an extension to the brain imaging data structure for electroencephalography. Scientific Data, 6, 103. https://doi.org/10.1038/s41597-019-0104-8\n\n---\nGenerated by MOABB 1.5.0 (Mother of All BCI Benchmarks)\nhttps://github.com/NeuroTechX/moabb\n","bids_version":"1.9.0","sessions_count":841,"publish_date":"2026-03-26 18:40:56","embedding_dirty":0,"license_tier":"sharealike","zarr_status":"ready","zarr_converted_at":"2026-09-04 08:40:39","zarr_store_count":1060,"zarr_index_etag":"bece801ca7fa543bd942f0ad9d43f9d0","zarr_source_commit":"fc8338165cc51046716b7c034b5bba74605a1f82","archive_status":"ready","archive_size":2875618749,"archive_retry_count":0,"records_status":"ready","archive_skip_reason":null,"zarr_errors":0,"zarr_failure_count":0,"zarr_deterministic":0,"zarr_failed_at":null,"num_dataset_citations":0,"num_datapaper_citations":2,"n_channels":31,"electrode_system":"10-10","has_hed":1,"hed_version":"8.4.0","is_exemplar":0,"bytes_present":3936025756,"data_complete":1,"withdrawn_at":null,"withdrawn_reason":null,"archive_complete":1,"archive_absent_files":0,"archive_declared_files":12731,"zarr_pool_breaks":0,"total_recording_duration":35258,"recording_duration_min":13,"recording_duration_max":63,"recording_count":1060,"recordings_unavailable":0,"recordings_measured":1060,"channel_count_min":37,"channel_count_max":37,"sampling_frequency":1000,"power_line_frequency":50,"eeg_reference":"nose","placement_scheme":"10-20 system","sweep_stamps":"{\"enrichment_updated_at\":\"2026-08-18 18:17:50\",\"metadata_updated_at\":\"2026-08-18 18:20:45\",\"archive_checked_at\":\"2026-08-18 18:24:44\",\"zarr_checked_at\":\"2026-06-07 17:58:37\",\"records_checked_at\":\"2026-08-18 18:24:50\",\"citations_updated_at\":\"2026-09-07 03:00:51\",\"channel_montage_checked_at\":\"2026-06-28 23:02:10\",\"hed_checked_at\":\"2026-06-30 04:32:26\",\"data_checked_at\":null,\"availability_report_at\":\"2026-08-22 03:01:21\",\"signal_defaults_at\":\"2026-09-02 11:50:10\",\"recording_stats_at\":\"2026-09-05 03:01:59\"}","participants":13,"num_citations":2,"latest_version":"v1.0.4","zarr_verify_status":null,"zarr_verified_at":null,"owner_username":"bruaristimunha","owner_github":"bruAristimunha","file_size_formatted":"3.68 GB","zarr_data_failures":null,"zarr_index_url":"https://zarr.nemar.org/nm000266/zarr/index.json","attestation_deposit_type":null,"attestation_key_status":null,"attestation_deidentified":null,"attestation_no_duplicate":null,"attestation_upstream_source":null,"attestation_accepted_at":null}}