{"success":true,"database":"eegdash","data":{"_id":"69d16e04897a7725c66f4c4d","dataset_id":"ds007477","associated_paper_doi":null,"authors":["Niu，Haijing","Zheng, Sha","Yuan, Haodong"],"bids_version":"1.7.0","contact_info":["浩东 苑"],"contributing_labs":null,"data_processed":true,"dataset_doi":"doi:10.18112/openneuro.ds007477.v1.0.1","datatypes":["fnirs"],"demographics":{"subjects_count":18,"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://openneuro.org/datasets/ds007477","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"945859376c4274cf5fadfd937392c24afba6ea144338ceeda9de5bd6b3613ee7","license":"CC0","n_contributing_labs":null,"name":"TimeSeries BIDS converted","readme":"This dataset was converted from TimeSeriesHbORT_18sub_twoSessionICAdenoise(1).mat using `convert_mat_to_bids.py`.\nNotes:\n- Review and confirm `*_nirs.json` (SamplingFrequency, NIRSChannelCount, source/detector mapping) before public release.\n- This README is a placeholder to satisfy BIDS recommendations; replace with dataset-specific information as needed.","recording_modality":["fnirs"],"senior_author":"Yuan, Haodong","sessions":["01","02"],"size_bytes":9384,"source":"openneuro","storage":{"backend":"s3","base":"s3://openneuro.org/ds007477","raw_key":"dataset_description.json","dep_keys":["CHANGES","README","participants.tsv"]},"study_design":null,"study_domain":null,"tasks":["ort"],"timestamps":{"digested_at":"2026-04-22T12:30:20.935047+00:00","dataset_created_at":"2026-03-06T03:08:35.425Z","dataset_modified_at":"2026-03-18T13:20:02.000Z"},"total_files":36,"computed_title":"TimeSeries BIDS converted","nchans_counts":[{"val":1,"count":36}],"sfreq_counts":[{"val":10.0,"count":36}],"stats_computed_at":"2026-04-22T23:16:00.312836+00:00","total_duration_s":null,"tagger_meta":{"config_hash":"3557b68bca409f28","metadata_hash":"21e0fbc243ea02bb","model":"openai/gpt-5.2","tagged_at":"2026-04-07T09:32:40.872789+00:00"},"tags":{"pathology":["Unknown"],"modality":["Other"],"type":["Unknown"],"confidence":{"pathology":0.4,"modality":0.7,"type":0.4},"reasoning":{"few_shot_analysis":"No close few-shot match by paradigm or population: the provided few-shot examples mostly describe explicit EEG paradigms (oddball, resting-state, sleep, motor imagery, reinforcement learning) and/or explicit clinical recruitment (e.g., Parkinson’s, TBI, epilepsy). This dataset metadata does not specify a recognizable EEG task paradigm, stimulus modality, or clinical population, so the few-shot conventions mainly guide the fallback behavior: when explicit facts are absent, prefer \"Unknown\" (or \"Other\" only when a concrete non-listed modality is explicitly indicated).","metadata_analysis":"Key available facts are minimal and largely about file conversion rather than experiment.\nQuoted snippets:\n1) \"This dataset was converted from TimeSeriesHbORT_18sub_twoSessionICAdenoise(1).mat\" (conversion provenance; no paradigm/population).\n2) \"Review and confirm `*_nirs.json` ... before public release.\" (explicit indication the BIDS modality is NIRS rather than EEG).\n3) \"participants_overview\": \"Subjects: 18\" (sample size only; no diagnosis/age/controls).\n4) \"tasks\": [\"ort\"] (task label present but undefined; no task description or stimuli).","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology:\n- Metadata says: only \"Subjects: 18\" with no diagnosis or recruitment criteria.\n- Few-shot pattern suggests: when diagnosis is explicitly stated, use that; otherwise do not assume Healthy.\n- Alignment: N/A (no pathology facts available) → choose Unknown.\n\nModality:\n- Metadata says: \"confirm `*_nirs.json`\" and the source file name includes \"Hb\" (\"TimeSeriesHbORT...\") which strongly suggests fNIRS/hemoglobin time series rather than a classic EEG stimulus modality.\n- Few-shot pattern suggests: map based on stimulus/input channel, but if the recording modality itself is outside the allowed list (NIRS), use \"Other\" rather than forcing Visual/Auditory/etc.\n- Alignment: aligns with using \"Other\" because NIRS is explicitly implied but not an allowed modality label.\n\nType:\n- Metadata says: only task code \"ort\" with no description of cognitive construct (no resting/sleep/motor/learning/etc.).\n- Few-shot pattern suggests: only assign a cognitive Type when the construct is described (e.g., \"digit span\" -> Memory; \"eyes closed resting\" -> Resting-state; \"motor imagery\" -> Motor).\n- Alignment: N/A (insufficient facts) → choose Unknown.","decision_summary":"Top-2 candidates and selection:\n\nPathology:\n- Candidate 1: Unknown — supported by absence of any recruitment/diagnosis text (\"Subjects: 18\" only).\n- Candidate 2: Healthy — weakly plausible because many small methodological datasets are healthy, but not stated anywhere.\nWinner: Unknown. Evidence alignment: no explicit pathology facts.\n\nModality:\n- Candidate 1: Other — supported by explicit reference to NIRS sidecar (\"confirm `*_nirs.json`\") and Hb time-series filename (\"TimeSeriesHbORT...\").\n- Candidate 2: Unknown — if we refuse to infer NIRS from these hints.\nWinner: Other. Evidence alignment: metadata explicitly points to NIRS, which is outside the allowed modality list.\n\nType:\n- Candidate 1: Unknown — task \"ort\" is undefined (\"tasks\": [\"ort\"]).\n- Candidate 2: Other — if we had any hint of a construct, but none is provided.\nWinner: Unknown. Evidence alignment: insufficient task/paradigm description.\n\nConfidence justification (quote-based):\n- Pathology low because only \"Subjects: 18\" is provided and no clinical keywords.\n- Modality moderate because \"*_nirs.json\" is explicitly mentioned and Hb appears in the source filename.\n- Type low because only an opaque task label \"ort\" is given with no description."}},"author_year":"Niu2026","canonical_name":null}}