{"success":true,"database":"eegdash","data":{"_id":"69d16e06897a7725c66f4ce5","dataset_id":"nm000347","associated_paper_doi":null,"authors":["Jian Shi","Danyang Chen","Xingwei Zhao","Zhixian Zhao","Shengjie Li","Yeguang Xu","Tao Ding","Zheng Zhu","Peng Zhang","Qing Ye","Yingxin Tang","Ping Zhang","Bo Tao","Zhouping Tang"],"bids_version":"1.9.0","contact_info":null,"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.1038/s41597-025-06100-7","datatypes":["eeg"],"demographics":{"subjects_count":37,"ages":[],"age_min":null,"age_max":null,"age_mean":null,"species":null,"sex_distribution":null,"handedness_distribution":{"r":37}},"experimental_modalities":null,"external_links":{"source_url":"https://openneuro.org/datasets/nm000347","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"4535cca1af5048d1d3fb6a3ea6b479bdf48169133a080ea292835611e0688ab8","license":"CC-BY-NC-ND-4.0","n_contributing_labs":null,"name":"Shi et al. 2025 — HEFMI-ICH: a hybrid EEG-fNIRS motor imagery dataset for brain-computer interface in intracerebral hemorrhage","readme":"HefmiIch2025\n============\nHybrid EEG-fNIRS MI dataset for ICH from Shi et al 2025.\nDataset Overview\n----------------\n  Code: HefmiIch2025\n  Paradigm: imagery\n  DOI: 10.1038/s41597-025-06100-7\n  Subjects: 37\n  Sessions per subject: 3\n  Events: left_hand=1, right_hand=2\n  Trial interval: [0, 10] s\n  File format: MAT (pre-epoched)\n  Data preprocessed: True\nAcquisition\n-----------\n  Sampling rate: 256.0 Hz\n  Number of channels: 32\n  Channel types: eeg=32\n  Channel names: FC1, AF3, AF4, CP1, CP2, CP6, Cz, C3, C4, T7, T8, FC2, FC5, FC6, Pz, CP5, PO3, PO4, Oz, Fp2, Fp1, Fz, F3, F4, F7, F8, P3, P4, P7, P8, O1, O2\n  Montage: biosemi32\n  Hardware: g.HIamp (g.tec medical engineering GmbH)\n  Line frequency: 50.0 Hz\n  Online filters: {}\nParticipants\n------------\n  Number of subjects: 37\n  Health status: mixed (17 healthy, 20 ICH patients)\n  Clinical population: intracerebral hemorrhage (ICH)\n  Age: min=20.0, max=65.0\n  Gender distribution: female=8, male=29\n  Handedness: right-handed\n  Species: human\nExperimental Protocol\n---------------------\n  Paradigm: imagery\n  Number of classes: 2\n  Class labels: left_hand, right_hand\n  Trial duration: 27.0 s\n  Study design: 2-class hand MI (left/right grasping) for ICH rehabilitation. 17 healthy + 20 ICH patients, 1-6 sessions per subject.\n  Feedback type: none\n  Stimulus type: directional arrow + auditory beep\n  Stimulus modalities: visual, auditory\n  Primary modality: visual\n  Synchronicity: synchronous\n  Mode: offline\nHED Event Annotations\n---------------------\n  Schema: HED 8.4.0 | Browse: https://www.hedtags.org/hed-schema-browser\n  left_hand\n    ├─ Sensory-event, Experimental-stimulus, Visual-presentation\n    └─ Agent-action\n       └─ Imagine\n          ├─ Move\n          └─ Left, Hand\n  right_hand\n    ├─ Sensory-event, Experimental-stimulus, Visual-presentation\n    └─ Agent-action\n       └─ Imagine\n          ├─ Move\n          └─ Right, Hand\nParadigm-Specific Parameters\n----------------------------\n  Detected paradigm: motor_imagery\n  Imagery tasks: left_hand, right_hand\n  Cue duration: 2.0 s\n  Imagery duration: 10.0 s\nData Structure\n--------------\n  Trials: 3330\n  Trials context: 37 subjects x ~3 sessions x 30 trials = ~3330\nSignal Processing\n-----------------\n  Classifiers: CSP+SVM, FBCSP+SVM, EEGBaseNet, TF+SVM\n  Feature extraction: CSP, FBCSP, time-frequency features\n  Frequency bands: preprocessing=[0.5, 30.0] Hz\n  Spatial filters: CSP, FBCSP\nCross-Validation\n----------------\n  Method: 5-fold\n  Folds: 5\n  Evaluation type: within_subject\nBCI Application\n---------------\n  Applications: rehabilitation\n  Environment: clinical\n  Online feedback: False\nTags\n----\n  Pathology: Healthy, Stroke\n  Modality: Motor\n  Type: Clinical, Research\nDocumentation\n-------------\n  DOI: 10.1038/s41597-025-06100-7\n  License: CC-BY-NC-ND-4.0\n  Investigators: Jian Shi, Danyang Chen, Xingwei Zhao, Zhixian Zhao, Shengjie Li, Yeguang Xu, Tao Ding, Zheng Zhu, Peng Zhang, Qing Ye, Yingxin Tang, Ping Zhang, Bo Tao, Zhouping Tang\n  Institution: Huazhong University of Science and Technology\n  Country: CN\n  Data URL: https://figshare.com/articles/dataset/28955456\n  Publication year: 2025\nReferences\n----------\nShi, J., Chen, D., et al. (2025). HEFMI-ICH: a hybrid EEG-fNIRS motor imagery dataset for brain-computer interface in intracerebral hemorrhage. Scientific Data. https://doi.org/10.1038/s41597-025-06100-7\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\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---\nGenerated by MOABB 1.5.0 (Mother of All BCI Benchmarks)\nhttps://github.com/NeuroTechX/moabb","recording_modality":["eeg"],"senior_author":null,"sessions":["0","1","2","3","4","5"],"size_bytes":2766104316,"source":"nemar","storage":{"backend":"s3","base":"s3://openneuro.org/nm000347","raw_key":"dataset_description.json","dep_keys":["README","participants.json","participants.tsv"]},"study_design":null,"study_domain":null,"tasks":["imagery"],"timestamps":{"digested_at":"2026-04-22T12:52:30.331066+00:00","dataset_created_at":null,"dataset_modified_at":null},"total_files":98,"computed_title":"Shi et al. 2025 — HEFMI-ICH: a hybrid EEG-fNIRS motor imagery dataset for brain-computer interface in intracerebral hemorrhage","nchans_counts":[{"val":32,"count":98}],"sfreq_counts":[{"val":256.0,"count":98}],"stats_computed_at":"2026-04-22T23:16:00.314715+00:00","total_duration_s":112307.6171875,"tagger_meta":{"config_hash":"3557b68bca409f28","metadata_hash":"81f66155974dc79a","model":"openai/gpt-5.2","tagged_at":"2026-04-07T09:32:40.872789+00:00"},"tags":{"pathology":["Other"],"modality":["Multisensory"],"type":["Motor"],"confidence":{"pathology":0.75,"modality":0.85,"type":0.8},"reasoning":{"few_shot_analysis":"Most similar few-shot by paradigm is the \"EEG Motor Movement/Imagery Dataset\" (motor/imagery runs) which is labeled Modality=Visual and Type=Motor; it sets the convention that motor imagery datasets are typically Type=Motor even if cues are visual. For multisensory cueing, the \"Cross-modal Oddball Task\" example uses simultaneous visual+auditory pre-cues and is labeled Modality=Multisensory, guiding the choice to prefer Multisensory when both visual and auditory stimuli are explicitly part of the cueing. For clinical recruitment conventions, multiple few-shots (e.g., Parkinson’s and Epilepsy datasets) show that when the metadata explicitly names a patient group, Pathology should reflect that clinical recruitment; here, the diagnosis is intracerebral hemorrhage (stroke), which is not an allowed pathology label, so the closest allowed label is Other.","metadata_analysis":"Key population facts: \"Health status: mixed (17 healthy, 20 ICH patients)\", and \"Clinical population: intracerebral hemorrhage (ICH)\", plus \"Study design: 2-class hand MI (left/right grasping) for ICH rehabilitation. 17 healthy + 20 ICH patients\". Key stimulus/modality facts: \"Stimulus type: directional arrow + auditory beep\", \"Stimulus modalities: visual, auditory\", and \"Primary modality: visual\". Key task/type facts: \"Paradigm: imagery\", \"Detected paradigm: motor_imagery\", \"Imagery tasks: left_hand, right_hand\", and \"2-class hand MI (left/right grasping)\".","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology — Metadata says: \"20 ICH patients\" and \"Clinical population: intracerebral hemorrhage (ICH)\". Few-shot pattern suggests: when a named clinical population is recruited, use a corresponding pathology label (e.g., Parkinson's, Epilepsy). ALIGN/CONFLICT: Aligns on being a clinical cohort, but CONFLICT with label availability (no Stroke/ICH label in allowed list). Resolution: use \"Other\" as the closest allowed label.\n\nModality — Metadata says: \"Stimulus modalities: visual, auditory\" and \"directional arrow + auditory beep\" (also \"Primary modality: visual\"). Few-shot pattern suggests: simultaneous visual+auditory cues map to \"Multisensory\" (as in Cross-modal Oddball). ALIGN/CONFLICT: Aligns; although visual is primary, explicit dual-modality cueing supports Multisensory.\n\nType — Metadata says: \"Detected paradigm: motor_imagery\" and \"2-class hand MI (left/right grasping)\" and events are left_hand/right_hand imagery. Few-shot pattern suggests: motor imagery datasets map to Type=\"Motor\" (as in EEG Motor Movement/Imagery Dataset). ALIGN/CONFLICT: Aligns. A secondary pull toward \"Clinical/Intervention\" exists because metadata mentions \"ICH rehabilitation\", but the core construct/task is motor imagery classification.","decision_summary":"Pathology top-2: (1) Other — supported by explicit recruitment of a non-listed clinical diagnosis: \"Clinical population: intracerebral hemorrhage (ICH)\", \"Health status: mixed (17 healthy, 20 ICH patients)\", and \"for ICH rehabilitation\"; label set lacks Stroke/ICH so Other is the best fit. (2) Healthy — plausible only because dataset includes \"17 healthy\" controls, but not the recruited patient focus. Evidence alignment: clinical-population fact aligns with few-shot clinical recruitment convention; final=Other.\n\nModality top-2: (1) Multisensory — supported by \"Stimulus modalities: visual, auditory\" and \"directional arrow + auditory beep\"; matches few-shot convention where combined visual+auditory cueing is Multisensory. (2) Visual — supported by \"Primary modality: visual\" and the arrow cue; but it ignores explicit auditory cueing. Evidence alignment: aligns; final=Multisensory.\n\nType top-2: (1) Motor — supported by \"Detected paradigm: motor_imagery\", \"2-class hand MI (left/right grasping)\", and imagery events (left_hand/right_hand); matches few-shot motor imagery convention (EEG Motor Movement/Imagery Dataset labeled Type=Motor). (2) Clinical/Intervention — supported by \"for ICH rehabilitation\" and presence of patients, but no actual intervention/feedback is described (\"Feedback type: none\"). Evidence alignment: aligns more strongly with Motor; final=Motor.\n\nConfidence basis: Pathology has multiple explicit clinical-population quotes but maps to a catch-all label due to allowed-label limitations; Modality has 3 explicit stimulus-modality quotes and strong few-shot analog; Type has multiple explicit motor-imagery quotes and strong few-shot analog, with some ambiguity due to rehab framing."}},"canonical_name":null,"name_confidence":0.72,"name_meta":{"suggested_at":"2026-04-14T10:18:35.344Z","model":"openai/gpt-5.2 + openai/gpt-5.4-mini + deterministic_fallback"},"name_source":"canonical","author_year":"HefmiIch2025"}}