{"success":true,"database":"eegdash","data":{"_id":"6953f4249276ef1ee07a347b","dataset_id":"ds007020","associated_paper_doi":null,"authors":["Simin Jamshidi","Arturo Espinoza","Soura Dasgupta","Nandakumar Narayanan"],"bids_version":"1.8.0","contact_info":["Simin Jamshidi"],"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.18112/openneuro.ds007020.v1.0.0","datatypes":["eeg"],"demographics":{"subjects_count":94,"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/ds007020","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"212c9a816896882c652db04dce9ecd1aff1db73274228bbe6eb8db3afd314002","license":"CC0","n_contributing_labs":null,"name":"EEG Mortality Dataset in Parkinson's Disease","readme":"This dataset contains de-identified resting-state EEG recordings from individuals with Parkinson’s disease (PD) and age-matched healthy control subjects. All EEG data were recorded using standard clinical EEG systems at Neurology Clinic.\nDataset Purpose:\nThis dataset was originally used to evaluate whether resting-state EEG can help distinguish subjects who were later deceased from those who remained living (mortality classification). Only de-identified EEG data and mortality labels are included.\nParticipant Information:\n- Participants are labeled as either \"living\" or \"deceased\" in participants.tsv\n- No other demographic or clinical information (age, cognition, UPDRS, disease duration, etc.) is included per data-sharing guidelines.\n- All participant IDs are anonymized following BIDS convention (e.g., sub-PD1301).\nEEG Acquisition Details:\n- Recording type: Resting-state EEG (eyes open)\n- Device: Clinical BrainVision EEG system\n- File formats: .vhdr, .eeg, .vmrk\n- Sampling rate: 500 Hz\n- Montage: Standard 10–20 international system\n- Recording condition: “task-rest” (no task)\nData Organization:\nData are structured following the BIDS (Brain Imaging Data Structure) EEG standard:\n    sub-<ID>/\n        ses-01/\n            eeg/\n                sub-<ID>_ses-01_task-rest_eeg.vhdr\n                sub-<ID>_ses-01_task-rest_eeg.eeg\n                sub-<ID>_ses-01_task-rest_eeg.vmrk\nMortality Label Format:\n- Living subjects: survival_status = \"living\"\n- Deceased subjects: survival_status = \"deceased\"\nEthics & Privacy:\nAll subjects provided consent for EEG recording at the University of Iowa Hospitals and Clinics. The publicly shared version here is fully de-identified and contains no clinical or personal health information other than mortality classification.\nSuggested Use:\nThis dataset can be used to explore EEG biomarkers of mortality risk, EEG signal characteristics in PD, or to build machine learning models for classification.\nQuestions or requests:\nPlease contact nandakumar-narayanan@uiowa.edu.","recording_modality":["eeg"],"senior_author":"Nandakumar Narayanan","sessions":["01"],"size_bytes":1870056562,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["rest"],"timestamps":{"digested_at":"2026-04-22T12:29:56.964784+00:00","dataset_created_at":"2025-12-02T04:13:14.727Z","dataset_modified_at":"2025-12-02T04:32:19.000Z"},"total_files":94,"storage":{"backend":"s3","base":"s3://openneuro.org/ds007020","raw_key":"dataset_description.json","dep_keys":["CHANGES","README.md","participants.json","participants.tsv"]},"tagger_meta":{"config_hash":"4a051be509a0e3d0","metadata_hash":"7e92996799e2babc","model":"openai/gpt-5.2","tagged_at":"2026-01-20T19:16:49.892819+00:00"},"tags":{"pathology":["Parkinson's"],"modality":["Resting State"],"type":["Clinical/Intervention"],"confidence":{"pathology":0.85,"modality":0.85,"type":0.8},"reasoning":{"few_shot_analysis":"Most similar few-shot conventions:\n- The dementia resting EEG example labels Modality as \"Resting State\" and Type as \"Clinical/Intervention\" when resting EEG is used in a clinical-discrimination/biomarker context (\"Alzheimer’s disease...resting state-closed eyes recordings\"). This guides mapping of a clinical resting-state biomarker dataset to Type=\"Clinical/Intervention\" rather than Type=\"Resting-state\".\n- The Parkinson’s cross-modal oddball example labels Pathology as \"Parkinson's\" when PD patients are explicitly recruited alongside controls. This guides Pathology labeling here because the metadata explicitly states a PD cohort plus age-matched controls.","metadata_analysis":"Key metadata facts (quotes):\n- Population/diagnosis: \"resting-state EEG recordings from individuals with Parkinson’s disease (PD) and age-matched healthy control subjects\"\n- Recording paradigm: \"Recording type: Resting-state EEG (eyes open)\" and \"Recording condition: “task-rest” (no task)\"\n- Study purpose (clinical/biomarker classification): \"evaluate whether resting-state EEG can help distinguish subjects who were later deceased from those who remained living (mortality classification)\" and \"Only de-identified EEG data and mortality labels are included.\"","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology:\n1) Metadata says: \"individuals with Parkinson’s disease (PD) and age-matched healthy control subjects\".\n2) Few-shot pattern suggests: PD-recruited cohorts map to Pathology=\"Parkinson's\" (see Parkinson oddball / reinforcement learning examples).\n3) ALIGN.\n\nModality:\n1) Metadata says: \"Recording type: Resting-state EEG (eyes open)\" and \"task-rest (no task)\".\n2) Few-shot pattern suggests: resting EEG maps to Modality=\"Resting State\" (see dementia resting EEG; sleep deprivation resting EEG).\n3) ALIGN.\n\nType:\n1) Metadata says the purpose is clinical outcome discrimination: \"mortality classification\" / \"EEG biomarkers of mortality risk\".\n2) Few-shot pattern suggests: clinical resting EEG used for diagnosis/biomarkers maps to Type=\"Clinical/Intervention\" (see dementia resting EEG labeled Clinical/Intervention), whereas non-clinical resting paradigms map to Type=\"Resting-state\" (see sleep deprivation example).\n3) Mostly ALIGN toward \"Clinical/Intervention\" because this dataset is explicitly framed as a clinical prognostic classification use-case.","decision_summary":"Top-2 candidate labels per category and final choice:\n\nPathology:\n- Candidate 1: Parkinson's — Supported by: \"individuals with Parkinson’s disease (PD)\".\n- Candidate 2: Healthy — Some controls included (\"age-matched healthy control subjects\"), but recruitment clearly includes PD.\nHead-to-head: Parkinson's wins because PD is an explicit recruited clinical cohort.\n\nModality:\n- Candidate 1: Resting State — Supported by: \"Resting-state EEG (eyes open)\" and \"task-rest (no task)\".\n- Candidate 2: Unknown — Only if resting-state were unclear, but it is explicit.\nHead-to-head: Resting State wins.\n\nType:\n- Candidate 1: Clinical/Intervention — Supported by: \"mortality classification\" and \"EEG biomarkers of mortality risk\" in a PD cohort.\n- Candidate 2: Resting-state — Task is resting and could be categorized by paradigm, but the stated primary aim is a clinical/prognostic classifier.\nHead-to-head: Clinical/Intervention wins because the dataset purpose is explicitly clinical outcome discrimination using EEG."}},"computed_title":"EEG Mortality Dataset in Parkinson's Disease","nchans_counts":[{"val":63,"count":76},{"val":64,"count":18}],"sfreq_counts":[{"val":500.0,"count":94}],"stats_computed_at":"2026-04-22T23:16:00.312241+00:00","total_duration_s":14783.066,"author_year":"Jamshidi2025","canonical_name":null}}