{"success":true,"database":"eegdash","data":{"_id":"6953f4249276ef1ee07a3461","dataset_id":"ds006695","associated_paper_doi":null,"authors":["Julie Onton","Sarah Mednick"],"bids_version":"1.8.0","contact_info":["Arnaud Delorme"],"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.18112/openneuro.ds006695.v1.0.2","datatypes":["eeg"],"demographics":{"subjects_count":19,"ages":[29,24,21,25,20,20,20,19,26,20,26,24,27,21,27,23,20,20,23],"age_min":19,"age_max":29,"age_mean":22.894736842105264,"species":null,"sex_distribution":{"f":10,"m":9},"handedness_distribution":null},"experimental_modalities":null,"external_links":{"source_url":"https://openneuro.org/datasets/ds006695","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"6cab129e095abf23c173508a11061b9f49e61685df9a7b1a0dc7441635bc3395","license":"CC0","n_contributing_labs":null,"name":"Validation of Sleep Staging with Forehead EEG Patch","readme":"# UCSD Forehead Patch Sleep Validation Dataset\nCurated EEG recordings for validating sleep staging from a three-electrode forehead patch against standard 33-channel polysomnography.\n## What is included\n- CGX patch `.set` files that contain `EEG.VisualHypnogram` and `EEG.SpectralScoring`. The 33-channel data is not included in this release -- this release only includes the three-electrode data. The 33-channel data will be released separately.\n## Sleep stage labels\n`EEG.VisualHypnogram` is manual scoring in 30-second epochs using the following integers\n1 equals Wake\n2 equals REM\n3 equals N1\n4 equals N2\n5 equals N3\n0 equals unknown or movement\n`EEG.SpectralScoring` is spectral staging from the forehead patch. One row per patch channel. One column per 30-second epoch (see publication).\n## Alignment policy\nThe 33-channel cap data used to score polysomnography and the 3-channel patch EEG data do not always start and stop at the same clock times. CGX patch data were aligned to the cap start time based on a spreadsheet completed by the data collector, so the start may be off by a few seconds. The 3-channel EEG data were segmented into 30-second windows, and the number of these windows should approximately match the number of values in the EEG.VisualHypnogram for the same dataset. If the patch data ended up shorter than the visual hypnogram, the hypnogram was trimmed at the end to match the patch length. If the hypnogram was longer, it was left untrimmed. In general, the mismatch at the end of the recording is less than one 30-second window.\n## Subject exclusions\n113 and 121 are excluded. The CGX patch was inadequate or unavailable.\n## Citation\nOnton JA, Simon KC, Morehouse AB, Shuster AE, Zhang J, Peña AA, Mednick SC. Validation of spectral sleep scoring with polysomnography using forehead EEG device. Frontiers in Sleep. 2024. doi 10.3389/frsle.2024.1349537.\nAmerican Academy of Sleep Medicine. The AASM manual for the scoring of sleep and associated events. 2007 and later.","recording_modality":["eeg"],"senior_author":"Sarah Mednick","sessions":[],"size_bytes":10052862715,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["sleep"],"timestamps":{"digested_at":"2026-04-22T12:29:34.728139+00:00","dataset_created_at":"2025-09-20T00:12:16.076Z","dataset_modified_at":"2025-09-20T16:30:31.000Z"},"total_files":19,"storage":{"backend":"s3","base":"s3://openneuro.org/ds006695","raw_key":"dataset_description.json","dep_keys":["CHANGES","README","participants.json","participants.tsv"]},"tagger_meta":{"config_hash":"3557b68bca409f28","metadata_hash":"7949c69829c2cc91","model":"openai/gpt-5.2","tagged_at":"2026-04-07T09:32:40.872789+00:00"},"tags":{"pathology":["Healthy"],"modality":["Sleep"],"type":["Sleep"],"confidence":{"pathology":0.7,"modality":0.9,"type":0.9},"reasoning":{"few_shot_analysis":"Most similar few-shot example: “Surrey cEEGrid sleep data set” (Healthy / Sleep / Sleep). It shares the same core paradigm: overnight EEG recordings for sleep monitoring/staging with no cognitive task. That example guides mapping sleep-staging/PSG validation datasets to Modality=Sleep and Type=Sleep. Population-wise, the Surrey example is explicitly “healthy participants”; the current dataset does not explicitly state a disorder group, so by convention it maps to a normative/Healthy cohort unless a clinical recruitment criterion is stated.","metadata_analysis":"Key metadata indicating sleep-staging paradigm: (1) Title: “Validation of Sleep Staging with Forehead EEG Patch”. (2) README: “Curated EEG recordings for validating sleep staging from a three-electrode forehead patch against standard 33-channel polysomnography.” (3) README: “Sleep stage labels… manual scoring in 30-second epochs… Wake… REM… N1… N2… N3”. (4) Tasks field: “sleep”. Participant info suggests non-clinical young adults: “Subjects: 19… Age range: 19-29” with no mention of any diagnosis or patient group.","paper_abstract_analysis":"No paper abstract text was provided in the metadata (only a citation line to the 2024 Frontiers in Sleep article).","evidence_alignment_check":"Pathology: Metadata says nothing about a clinical diagnosis/recruitment (e.g., only “Subjects: 19… Age range: 19-29” and a methods-focused “validation… against… polysomnography”), while the few-shot pattern for similar sleep-staging datasets typically uses Healthy when no disorder group is specified. ALIGN (no conflict; absence of clinical group supports Healthy by convention).\n\nModality: Metadata explicitly says “sleep staging”, “polysomnography”, and provides sleep-stage labels (Wake/REM/N1/N2/N3) and task “sleep”. Few-shot sleep datasets map this to Modality=Sleep. ALIGN.\n\nType: Metadata focus is sleep staging/validation with hypnogram labels (“manual scoring in 30-second epochs… REM… N1… N2… N3”) rather than cognition. Few-shot convention labels these as Type=Sleep. ALIGN.","decision_summary":"Top-2 candidates per category with head-to-head selection:\n\nPathology:\n1) Healthy — Evidence: no clinical recruitment described; young adult sample “Age range: 19-29”; dataset framed as device/algorithm “Validation of Sleep Staging… against… polysomnography.” Few-shot sleep example uses Healthy when participants are not described as patients.\n2) Unknown — Evidence: metadata never explicitly states “healthy” or “control”.\nWinner: Healthy (stronger because the dataset is methodological validation without any stated disorder group).\n\nModality:\n1) Sleep — Evidence: “validating sleep staging”; “polysomnography”; “Sleep stage labels… Wake… REM… N1… N2… N3”; task listed as “sleep”.\n2) Resting State — Evidence: sleep involves no active task, but it is explicitly sleep rather than wakeful rest.\nWinner: Sleep.\n\nType:\n1) Sleep — Evidence: hypnogram/sleep-stage scoring focus (“manual scoring in 30-second epochs… REM… N1… N2… N3”; “validating sleep staging”).\n2) Clinical/Intervention — Evidence: ‘validation’ could be misconstrued as an intervention/clinical device study, but there is no patient treatment/intervention emphasis; it is primarily sleep staging.\nWinner: Sleep.\n\nConfidence justifications:\n- Pathology 0.7: supported by 1 clear contextual quote showing non-clinical validation framing and lack of diagnosis (“validating sleep staging… against… polysomnography”) plus participant demographics (“Age range: 19-29”), but no explicit “healthy” statement.\n- Modality 0.9: 3+ explicit sleep-related quotes/features (title/README sleep staging, PSG reference, sleep stage labels, task=sleep) + strong few-shot match.\n- Type 0.9: 3+ explicit quotes/features indicating sleep staging/hypnogram scoring + strong few-shot match."}},"computed_title":"Validation of Sleep Staging with Forehead EEG Patch","nchans_counts":[{"val":3,"count":19}],"sfreq_counts":[{"val":500.0,"count":19}],"stats_computed_at":"2026-04-22T23:16:00.311879+00:00","total_duration_s":591328.71,"canonical_name":null,"name_confidence":0.76,"name_meta":{"suggested_at":"2026-04-14T10:18:35.343Z","model":"openai/gpt-5.2 + openai/gpt-5.4-mini + deterministic_fallback"},"name_source":"author_year","author_year":"Onton2025"}}