| Title | Public EEG Dataset |
| Version | 23th May, 2023 |
ERP Core is an open source human ERP research project that covers paradigms and datasets commonly used in ERP research (6 ERP paradigms and 7 ERP components in total). One of the initiators of the project is Professor Steve Luck, whom we know well, and the main purpose of the project is to better advance ERP technology into social practice and to spread ERP technology for the benefit of humanity.
Project website: https://erpinfo.org/erp-core Project introduction: https://www.sciencedirect.com/science/article/pii/S1053811920309502 Access to the dataset: https://osf.io/thsqg/
OpenNeuro is a free and open source neuroimaging database sharing platform created by Poldrack and his team, providing a large number of MRI, MEG, EEG, iEEG, ECoG, ASL and PET datasets available for sharing. 540 publicly available As of today (May 2021), there are 540 publicly available datasets on OpenNeuro, and a total of 18,108 researchers have joined the platform to contribute to the database. Simply open OpenNeuro and search for relevant types of datasets by searching keywords (e.g., EEG) as needed, with no registration required. Today I am sharing with you an ERP dataset in OpenNeuro using the go / nogo detection and classification task. In this dataset, the authors have collected ERP data from 7 male and 7 female subjects each, and have also provided data analysis scripts and references.
OpenNeuro project website: https://openneuro.org/ OpenNeuro retrieval address: https://openneuro.org/public/datasets The Go/No_Go dataset is available at: https://openneuro.org/datasets/ds002680/versions/1.0.0
There are a few free and publicly available epilepsy datasets that can be collected on the internet, as many databases are part of the fee-based European Epilepsy Database (EED) project and need to be purchased separately.
The author has carefully searched and compiled the following small sample of epilepsy datasets as much as possible for your study and use.
EEG sample data from 10 patients with epilepsy
Data from epilepsy patients collected at the Neurology and Sleep Center in Hauz Khas, New Delhi, were based on a 10-20 distribution with a sampling frequency of 200 Hz in the MAT file. During the acquisition, the data were band-pass filtered in the range: 0.5 to 70 Hz and subdivided into pre-seizure, inter-seizure and seizure phases. Detailed information is available via txt text in the zip archive.
The dataset is available at: https://www.researchgate.net/publication/308719109_EEG_Epilepsy_Datasets
CHB-MIT scalp EEG database
The database is collected from Boston Children's Hospital and includes EEG records from pediatric patients with intractable seizures. Detailed information can be found on the database web page.
Access the database at: https://physionet.org/content/chbmit/1.0.0/
Kaggle competition on seizure prediction Dataset
This dataset is from the Kaggle Epilepsy Prediction Project competition. ECoG data from dogs and humans were collected, and detailed information can be found in the project website description.
Project description: https://www.kaggle.com/c/seizure-detection/overview Database access address: https://www.kaggle.com/c/seizure-detection/data
CHB-MIT Scalp EEG Database
The database contains EEG data from 23 pediatric epilepsy patients.
Database presentation literature: https://dspace.mit.edu/handle/1721.1/54669 Project and database address: https://archive.physionet.org/pn6/chbmit/
Resting-state EEG dataset of 41 Parkinson's patients
In the study conducted under this dataset, the researchers attempted to distinguish Parkinson's patients from normal individuals by EEG algorithms. The results were published in the journal Parkinsonism & Related Disorders in October 2020.
Published results: https://www.sciencedirect.com/science/article/pii/S1353802020306672?via%3Dihub Access to the dataset: https://bit.ly/3pP1pts
Task state EEG dataset for 26 Parkinson's patients
In the study conducted under this dataset, the investigators explored whether frontal theta and beta oscillations were abnormal in Parkinson's patients during Lower Limb Pedaling Task (LPT).
Published results: https://www.sciencedirect.com/science/article/abs/pii/S1388245720300092 Dataset available at: https://bit.ly/32dsmMS
Task state EEG data from 28 Parkinson's patients
In the study conducted under this dataset, the investigators aimed to investigate EEG markers for the clinical diagnosis of Parkinson's disease. The relevant findings were published in the journal Brain Research in January 2020.
Published results: https://www.sciencedirect.com/science/article/abs/pii/S0006899319305955?via%3Dihub Access to the dataset: https://bit.ly/2AIPl9b
Task state EEG data from 28 Parkinson's patients
In the study conducted under this dataset, the researchers found that patients with Parkinson's disease had reduced frontal mesolimbic theta activity during the performance of tasks related to cognitive control. Related findings were published in Neuropsychologia in 2018.
Published results: https://www.sciencedirect.com/science/article/abs/pii/S0028393218302185?via%3Dihub Access to dataset: http://bit.ly/2FauZTt
Resting-state EEG dataset of 27 Parkinson's patients
The dataset was resting-state EEG data from 27 Parkinson's patients, and in the study conducted under this dataset, the researchers developed a classification algorithm to classify healthy individuals and Parkinson's patients with a specificity of 82%. The related findings were published in Clinical Neurophysiology in 2018.
Published results: https://www.sciencedirect.com/science/article/abs/pii/S1388245717311719 Dataset access: http://bit.ly/2rfCkNP
In a study conducted under the Flanker Task, a dataset of task-state EEG data from 23 patients with OCD, researchers explored differences in the negative waves of error associations in OCD patients under a lateral inhibition task compared to healthy controls. The results were published in the journal Neuropsychologia in 2009.
Published results: https://www.sciencedirect.com/science/article/abs/pii/S0028393209001298?via%3Dihub Data set available at: https://bit.ly/2MG4ZHz
Task state depression EEG dataset
The dataset was task-state EEG data (Reinforcement Learning Task) from 46 depressed patients, and in the study conducted under this dataset, the researchers explored the differences in the negative waves of false associations in OCD patients under the lateral inhibition task compared to healthy controls. The related findings were published in 2019.
Published results: https://doi.org/10.1162/cpsy_a_00024
Access to the dataset: https://bit.ly/2F11Zwv
MODMA dataset
MODMA dataset is a professional open multimodal database for brain disorders and the website currently offers EEG and audio databases. The database currently provides MDD EEG data, as confirmed by the author. However, the dataset cannot be downloaded directly for access, and requires an account registration with an institutional email address and approval before it can be downloaded for use.
Database introduction: http://modma.lzu.edu.cn/data/index/ Database introduction literature: https://arxiv.org/pdf/2002.09283.pdf Data set access address: http://modma.lzu.edu.cn/data/application/#data_1
Note: Please make sure to read the account registration requirements at the top of the website carefully, otherwise it will fail to register.
Task state EEG data in 46 patients with schizophrenia
In a study conducted under this dataset of task-state EEG data (Cost Conflict Simon Task) from 46 patients with schizophrenia, researchers explored the presence of enhanced response conflict effects on spatial response in patients with schizophrenia. The related findings were published in Neuropsychologia in 2019.
Published results: https://www.sciencedirect.com/science/article/pii/S0028393218301726 The dataset is available at: https://bit.ly/2J7BeJc
Comparative data between normal and schizophrenic individuals
This dataset is EEG data for healthy adolescents and adolescents diagnosed with schizophrenia.
Details of the data and how to access them: http://brain.bio.msu.ru/eeg_schizophrenia.htm
Task state data for patients with schizophrenia
This dataset is the EEG data of schizophrenic patients performing basic sensory tasks.
Details of the data and how to access them: https://www.kaggle.com/broach/button-tone-sz
The BCI competition was created to provide high-quality neuroscience data to a wide range of BCI researchers. There are currently three datasets publicly available on the web, the 2nd, 3rd and 4th BCI datasets, which are available as follows:
The 2nd BCI competition dataset: https://www.bbci.de/competition/ii/ The 3rd BCI competition dataset: https://www.bbci.de/competition/iii/ The 4th BCI competition dataset: https://www.bbci.de/competition/iv/
Due to the COVID-19 epidemic, the 2020 International BCI Competition award event could not be held as scheduled. The organizing committee of the competition decided to share the tested data without open labels to the public, and users can use the competition test data for free.
Introduction to the 2020 BCI Competition wiki: https://osf.io/pq7vb/wiki/home/ 2020 BCI Competition dataset access: https://osf.io/pq7vb/
BNCI Horizon 2020 is a project funded by the European Commission Framework Programme 7. The project aims to make publicly available BCI datasets available to the general public to aid BCI research and development. Currently the project has aggregated 28 BCI-related datasets. Users can download and use them for free when they follow the license.
Project website: http://bnci-horizon-2020.eu/ Project introduction: http://bnci-horizon-2020.eu/project Database access address: http://bnci-horizon-2020.eu/database/data-sets
The MEDICON 2019 Scientific Challenge is a BCI program challenge in which participants can test BCI-based p300 components from 15 clinical EEG datasets to train adolescents with ASD to follow social cues in an experiment. Each subject received 7 training sessions over a 4-month period.
Project description and access to the dataset at: https://www.medicon2019.org/scientific-challenge/
Havard dataverse BCI motion imagery dataset, contributed by a researcher at Tianjin University.
Dataset is available: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/27306
Dataset is available: https://www.physionet.org/content/eegmmidb/1.0.0/
Dataset is available: https://www.physionet.org/content/erpbci/1.0.0/
To accelerate research on pediatric sleep and its link to health, National Children's Hospital (NCH) and Carnegie Mellon University (CMU) have launched the NCH Sleep Database. The dataset was conducted at NCH in Columbus, Ohio, USA between 2017 and 2019 on 3,984 pediatric sleep studies in 3,673 unique patients, as well as longitudinal clinical data on patients. Published polysomnography (PSG) monitors contain physiological signals from patients as well as technician assessments of sleep stages and descriptions of other abnormalities.
The dataset is available at: https://www.physionet.org/content/nch-sleep/0.1.0/
This dataset contains data from 197 sleep EEG cases and the data are in EDF format.
The database is available at: https://www.physionet.org/content/sleep-edfx/1.0.0/ EDFbrowser: https://www.teuniz.net/edfbrowser/
The dataset contains PSG data from 100 patients.
Project description: https://sleeptight.isr.uc.pt/ Database access address: https://sleeptight.isr.uc.pt/?page_id=48 Database introduction: http://dataset.isr.uc.pt/ISRUC_Sleep/Content.pdf Database documentation: https://www.researchgate.net/publication/283734463_ISRUC-Sleep_A_comprehensive_public_dataset_for_sleep_researchers
Differentiation of sleep stages using brief 30s EEG data collected with a simple headband EEG device(Dreem).
Project description: https://www.kaggle.com/c/dreem-sleep-stages/overview Database access: https://www.kaggle.com/c/dreem-sleep-stages/data
This dataset contains 1,110 sleep EEGs in MATLAB that were recorded in clinical conditions for 10 age groups of newborns from 36 to 45 weeks.
Project description literature: https://www.ncbi.nlm.nih.gov/pubmed/15055799 /
https://www.ncbi.nlm.nih.gov/pubmed/10406020
Database access: https://figshare.com/articles/dataset/Newborn_sleep_EEG_data/4729840
Published results: https://link.springer.com/article/10.1007/s11682-019-00171-y The dataset is available at: https://bit.ly/2GNej6D
Published results: https://www.sciencedirect.com/science/article/abs/pii/S0028393219301630?via%3Dihub The dataset is available at: https://bit.ly/2QkPB4d
HBND is a public interest organization providing a database of child and adolescent psychiatric research. The organization provides over 10,000+ children and adolescents with brain imaging data, including EEG, MRI, and other data. The project has currently released a total of 9 EEG database versions with approximately 3000 cases of data.
Project description literature: https://www.nature.com/articles/sdata2017181 Database access address: http://fcon_1000.projects.nitrc.org/indi/cmi_healthy_brain_network/sharing_neuro.html
Temple University EEG database with 12,000 patients, 16 channels, EEG data in EDF format.
Project address: https://www.isip.piconepress.com/projects/tuh_eeg/
Note: The database requires a registered account to access
A sentiment database containing video data and EEG data, accessible by registration invitation system, need to obtain authorization to access resources.
Project address: http://www.eecs.qmul.ac.uk/mmv/datasets/deap/
Note: When applying for authorization, be sure to read the authorization instructions carefully and fill out the authorization information form according to the requirements and format.
[TBC]