{"success":true,"database":"eegdash","data":{"_id":"6953f4249276ef1ee07a337b","dataset_id":"ds004706","associated_paper_doi":null,"authors":["Joseph H. Rudoler","Matthew R. Dougherty","Brandon S. Katerman","James P. Bruska","Woohyeuk Chang","David J. Halpern","Nicholas B. Diamond","Michael J. Kahana"],"bids_version":"1.6.0","contact_info":["Joseph Rudoler"],"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.18112/openneuro.ds004706.v1.0.0","datatypes":["eeg"],"demographics":{"subjects_count":34,"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/ds004706","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"d6c39e8513bf7420b1e9f05a4969ba7b0d3df43b3cd121241992983c923612b2","license":"CC0","n_contributing_labs":null,"name":"Spatial memory and non-invasive closed-loop stimulus timing","readme":"This dataset contains behavioral events and electrophysiological recordings from an experiment run in the Computational Memory Lab at the University of Pennsylvania from 2021-2022 with funding from U.S. Army Medical Research and Development Command (USAMRDC) through the Medical Technology Enterprise Consortium (MTEC) project MTEC-20-06-MOM-013, \"Restoring memory with task-independent semi-chronic closed-loop direct brain stimulation and non-invasive closed-loop stimulus timing optimization\". This experiment constitutes the non-invasive portion of the project, which targeted memory improvement through classifier-based stimulus presentation.\nThe experiment is a hybrid spatial-navigation and free recall paradigm in which subjects play the role of a courier delivering items to stores across a virtual town, and are subsequently asked to recall their deliveries. There are two phases - \"read-only\" and \"closed-loop\". In read-only sessions, there is no classifier-based timing manipulation and participants simply perform the task in order to generate training data for the models used in subsequent closed-loop sessions. After collecting sufficient training data, classifier models predict recall in closed-loop sessions and the stimulus presentation is timed to coincide with predicted good or bad memory encoding.\nTwo publications are based on this experiment:\n[\"Neural correlates of memory in an immersive spatiotemporal context\"](https://www.biorxiv.org/content/10.1101/2022.11.30.518606) studies the navigation and memory dynamics in read-only sessions, and \"Optimizing learning via real-time neural decoding\" (link pending) explores the results of the closed-loop manipulation.\nNote: memory dynamics in closed-loop sessions are potentially influenced by the closed-loop timing manipulation, and so may be biased in a way that precludes them from analyses of general mnemonic function. The read-only sessions, however, were not subject to this manipulation and therefore can be used for studying spatial and episodic memory (as in the first paper mentioned above).","recording_modality":["eeg"],"senior_author":"Michael J. Kahana","sessions":["0","1","2","3","4","5","6","7","8"],"size_bytes":1426144562699,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["NiclsCourierClosedLoop","NiclsCourierReadOnly"],"timestamps":{"digested_at":"2026-04-22T12:26:47.910143+00:00","dataset_created_at":"2023-08-16T20:56:12.789Z","dataset_modified_at":"2023-08-22T12:13:59.000Z"},"total_files":298,"storage":{"backend":"s3","base":"s3://openneuro.org/ds004706","raw_key":"dataset_description.json","dep_keys":["CHANGES","README","participants.json","participants.tsv"]},"tagger_meta":{"config_hash":"4a051be509a0e3d0","metadata_hash":"0287abf69b42dafe","model":"openai/gpt-5.2","tagged_at":"2026-01-20T10:45:46.412992+00:00"},"tags":{"pathology":["Healthy"],"modality":["Visual"],"type":["Memory"],"confidence":{"pathology":0.6,"modality":0.6,"type":0.8},"reasoning":{"few_shot_analysis":"Closest few-shot analog is the dataset titled \"EEG, pupillometry... digit span task\" (Healthy / Auditory / Memory). That example shows the convention that when the paradigm is explicitly a memory task (e.g., digit span with recall), the Type label is set to \"Memory\" (not \"Attention\" or \"Decision-making\"), and Pathology is \"Healthy\" when no clinical recruitment is described. Unlike that few-shot (auditory digit presentation), the current dataset’s task is an immersive virtual navigation + free recall paradigm, which by convention implies predominantly visual stimulation, so Modality should shift to \"Visual\" while keeping Type as \"Memory\".","metadata_analysis":"Key task and study-goal facts from the README:\n1) Memory focus: \"The experiment is a hybrid spatial-navigation and free recall paradigm... and are subsequently asked to recall their deliveries.\"\n2) Intervention/closed-loop timing for memory: \"targeted memory improvement through classifier-based stimulus presentation\" and \"stimulus presentation is timed to coincide with predicted good or bad memory encoding.\"\nNo explicit clinical population is mentioned anywhere in the provided metadata (no patient group, diagnosis, or condition-based recruitment described).","paper_abstract_analysis":"No useful paper information.","evidence_alignment_check":"Pathology:\n- Metadata says: no diagnosis/clinical group is stated; the README only describes funding/project goals and the task (e.g., \"memory improvement\"), but does not mention patients.\n- Few-shot pattern suggests: when participants are not described as a clinical cohort, label as Healthy.\n- ALIGNMENT: Align (no conflict).\n\nModality:\n- Metadata says: \"spatial-navigation... across a virtual town\" (implies visually presented virtual environment) and \"stimulus presentation\" (unspecified sensory channel).\n- Few-shot pattern suggests: modality follows stimulus channel; virtual navigation tasks are typically Visual when not otherwise specified.\n- ALIGNMENT: Mostly align, but metadata does not explicitly say \"visual\"; inference required.\n\nType:\n- Metadata says: \"free recall paradigm\" and repeated references to \"memory encoding\" and \"mnemonic function\".\n- Few-shot pattern suggests: explicit recall/encoding paradigms map to Type=Memory.\n- ALIGNMENT: Align strongly (no conflict).","decision_summary":"Top-2 candidates per category with head-to-head comparison:\n\nPathology:\n1) Healthy — Evidence: no clinical recruitment described; task-focused README (e.g., \"hybrid spatial-navigation and free recall paradigm\") without diagnosis terms. Matches few-shot convention where non-clinical cohorts are labeled Healthy.\n2) Unknown — Evidence: participant details are not provided explicitly (no subjects.tsv excerpt), so pathology cannot be confirmed from demographics.\nDecision: Healthy (stronger because absence of any clinical cohort description is consistent with Healthy labeling in the few-shots). Confidence is moderate due to missing explicit participant description.\n\nModality:\n1) Visual — Evidence: \"across a virtual town\" and immersive navigation strongly imply visually presented stimuli; no mention of auditory/tactile dominance.\n2) Multisensory — Evidence: immersive VR-like tasks can be multisensory in principle, but metadata does not mention audio/haptics.\nDecision: Visual (stronger because the only explicit stimulus context is a virtual town/navigation). Confidence moderate since \"visual\" is not explicitly stated.\n\nType:\n1) Memory — Evidence: \"free recall paradigm\"; \"predict recall\"; \"memory encoding\"; \"studying spatial and episodic memory\".\n2) Learning — Evidence: closed-loop \"stimulus timing optimization\" and \"Optimizing learning\" suggests a learning optimization angle, but the core measured construct is recall/encoding.\nDecision: Memory (stronger due to repeated explicit recall/episodic memory framing). Confidence fairly high given multiple direct memory-related phrases."}},"nemar_citation_count":3,"computed_title":"Spatial memory and non-invasive closed-loop stimulus timing","nchans_counts":[{"val":137,"count":298}],"sfreq_counts":[{"val":2048.0,"count":298}],"stats_computed_at":"2026-04-22T23:16:00.308193+00:00","source_url":"https://openneuro.org/datasets/ds004706","total_duration_s":1694181.8544921875,"author_year":"Rudoler2023","canonical_name":null}}