{"success":true,"database":"eegdash","data":{"_id":"69d16e05897a7725c66f4ccf","dataset_id":"nm000272","associated_paper_doi":null,"authors":["Michele Romani","Devis Zanoni","Elisabetta Farella","Luca Turchet"],"bids_version":"1.9.0","contact_info":null,"contributing_labs":null,"data_processed":false,"dataset_doi":"10.82901/nemar.nm000272","datatypes":["eeg"],"demographics":{"subjects_count":45,"ages":[],"age_min":null,"age_max":null,"age_mean":null,"species":null,"sex_distribution":null,"handedness_distribution":null},"experimental_modalities":null,"external_links":{"source_url":"https://nemar.org/dataexplorer/detail/nm000272","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"85130a45eec9dcaf68b76cb985eeec443ea2c2fedfaf18ec4540ff1637e66e27","license":"CC-BY-4.0","n_contributing_labs":null,"name":"Romani et al. 2025 — BrainForm: a Serious Game for BCI Training and Data Collection (P300 ERP, University of Trento)","readme":"[![DOI](https://img.shields.io/badge/DOI-10.82901%2Fnemar.nm000272-blue)](https://doi.org/10.82901/nemar.nm000272)\n# BrainForm: a P300 ERP EEG dataset from a serious game for BCI training and data collection\n## Summary\nThis dataset contains scalp electroencephalography (EEG) recordings collected with\n**BrainForm**, a gamified (serious-game) **P300 event-related potential (ERP)**\nbrain–computer interface (BCI) designed for scalable data collection using consumer\nhardware and a minimal setup. Participants completed repeated runs of a P300 spelling /\nselection task, enabling study of BCI skill acquisition across sessions and of the\nperceptual/performance effects of different visual stimulation textures.\nThe BIDS conversion in this NEMAR record contains EEG data for **22 participants**\n(`sub-*` folders), organised under the `eeg/` modality with a P300 task (`task-p300`).\n## Modality and paradigm\n- **Modality:** EEG (scalp electroencephalography)\n- **Task / paradigm:** P300 ERP oddball / speller within a serious game (`task-p300`)\n- **Focus:** BCI training, human factors, machine learning on ERP data\n## Participants\nThis BIDS dataset includes **22 subjects**. The original study used a within-subject\ndesign with multiple runs and two task complexities. Refer to `participants.tsv` and the\npaper for details.\n## Original dataset / data paper\nPlease cite the original paper when using this dataset:\n> Romani, M., Zanoni, D., Farella, E., & Turchet, L. (2025). *BrainForm: a Serious Game\n> for BCI Training and Data Collection.* arXiv:2510.10169.\n> https://doi.org/10.48550/arXiv.2510.10169\n- **DOI / preprint:** [arXiv:2510.10169](https://doi.org/10.48550/arXiv.2510.10169)\n- **Source / project:** University of Trento and Fondazione Bruno Kessler (FBK)\n- **Related record:** https://zenodo.org/records/17225966\n## Attribution\nAll data were collected by the original authors (Michele Romani and colleagues,\nUniversity of Trento / FBK). Please credit the original creators and cite the paper\nabove. This NEMAR record redistributes the dataset in BIDS format; EEG-BIDS and related\nBIDS tools were used only for standardisation, not as the source of the data.\n## License\nCC-BY-4.0 (see `dataset_description.json`).","recording_modality":["eeg"],"senior_author":"Luca Turchet","sessions":["0cb","0grain","1cb","1grain","2cbExtra","2grainExtra"],"size_bytes":248663467,"source":"nemar","storage":{"backend":"nemar","base":"s3://nemar/nm000272","raw_key":"dataset_description.json","dep_keys":["README.md","participants.tsv"]},"study_design":null,"study_domain":null,"tasks":["ERP","p300"],"timestamps":{"digested_at":"2026-10-22T08:31:47.858740+00:00","dataset_created_at":null,"dataset_modified_at":"2026-08-20 19:22:15"},"total_files":181,"computed_title":"Romani et al. 2025 — BrainForm: a Serious Game for BCI Training and Data Collection (P300 ERP, University of Trento)","nchans_counts":[{"val":8,"count":120}],"sfreq_counts":[{"val":250.0,"count":120}],"stats_computed_at":"2026-10-22T09:54:07.649986+00:00","total_duration_s":22602.0,"tagger_meta":{"model":"gpt-6-sol","taxonomy":"v2","config_hash":"39931253008fe2df","tagged_at":"2026-10-06T13:13:11Z","source":"eegdash-llm-tagger"},"tags":{"pathology":["Healthy"],"modality":["Visual"],"type":["Learning","Attention"],"confidence":{"pathology":0.6,"modality":0.8,"type":0.9},"reasoning":{"few_shot_analysis":"The Parkinson's cross-modal oddball example maps an oddball paradigm to Attention and uses stimulus channels, rather than EEG recording technology, for Modality. BrainForm's P300 oddball/speller is similarly stimulus-driven, but its metadata specifies visual stimulation. The healthy motor-imagery BCI example shows that BCI use alone does not determine Type; the stated research aim does.","metadata_analysis":"The readme describes a \"P300 spelling / selection task\" and a \"P300 ERP oddball / speller within a serious game.\" It says the runs enable study of \"BCI skill acquisition across sessions\" and \"perceptual/performance effects of different visual stimulation textures.\" It reports \"22 participants,\" while participants_overview says \"Subjects: 45\"; neither field identifies a clinical recruitment condition.","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology: Metadata says \"22 participants\" but names no diagnosis; the healthy-participant BCI example suggests Healthy. They align insofar as neither indicates clinical recruitment, though health is not explicitly confirmed. Modality: Metadata says \"visual stimulation textures\"; the oddball example suggests labeling the stimulus channel, not the EEG recording. They align on that convention, with Visual replacing the example's Multisensory because no auditory stimulus is stated here. Type: Metadata says \"P300 ERP oddball / speller\" and \"BCI skill acquisition across sessions\"; the oddball example suggests Attention, while the stated acquisition aim supports Learning. These align. No metadata fact is overridden by a few-shot pattern.","decision_summary":"Pathology—Healthy versus Unknown: no clinical condition is named, so Healthy is the stronger contextual inference, but confidence is limited (0.6). Modality—Visual versus Multisensory: \"visual stimulation textures\" directly supports Visual; no second stimulus channel is described (0.8). Type—Learning versus Attention: \"BCI training,\" \"BCI skill acquisition across sessions,\" and \"P300 ERP oddball / speller\" support Learning as the stated longitudinal aim and Attention as the oddball construct (0.9). Perception is less central than these two despite the stated visual-texture effects."}},"canonical_name":null,"name_confidence":0.42,"name_meta":{"suggested_at":"2026-04-14T10:18:35.344Z","model":"openai/gpt-5.2 + openai/gpt-5.4-mini + deterministic_fallback"},"name_source":"author_year","author_year":"Romani2025_BF_ERP","associated_paper_meta":{"channel":"nemar/IsDerivedFrom","confidence":"high","author_overlap":4,"is_oa":true,"oa_status":"preprint","source":"paper_resolver"},"acknowledgements":"University of Trento and Fondazione Bruno Kessler","references_and_links":["https://doi.org/10.48550/arXiv.2510.10169","https://zenodo.org/records/17225966"],"source_datasets":[{"DOI":"doi:10.48550/arXiv.2510.10169"}],"bad_channels_info":null,"n_records":120}}