| # Cavanagh2018b: EEG Parkinson's Classification Dataset in Resting-State | |
| The Cavanagh2018b dataset contains resting-state EEG recordings from the same group of people used in the Cavanagh2018a study. Participants included 28 individuals diagnosed with Parkinson's disease and 28 age- and sex-matched control participants. Each subject underwent a 2-minute resting-state EEG recording session with eyes open. These recordings were collected in a seated posture, prior to or following the auditory oddball task, under the same EEG setup (64-channel cap, 500 Hz sampling rate). | |
| This dataset serves as a complementary baseline condition for evaluating spontaneous brain dynamics in Parkinson's disease. Though not central to the novelty task paradigm, resting-state data may be useful for investigations into low-frequency oscillations or non-task-based classification approaches of Parkinson's disease. | |
| ## Paper | |
| Cavanagh, J. F., Kumar, P., Mueller, A. A., Richardson, S. P., & Mueen, A. (2018). **Diminished EEG habituation to novel events effectively classifies Parkinson's patients**. _Clinical Neurophysiology_, 129(2), 409-418. | |
| DISCLAIMER: We (DISCO) are NOT the owners or creators of this dataset, but we merely uploaded it here, to support our's ("EEG-Bench") and other's work on EEG benchmarking. | |
| ## Dataset Structure | |
| The `IMPORT_ME_REST.xlsx` file contains some information about the patients with PD (sex, age, medication status at first recording), as well as their matched control patients. | |
| ### Filename Format | |
| ``` | |
| [PID]_[SESSION]_PD_REST.mat | |
| ``` | |
| PID is the patient ID (e.g. `813`), while SESSION distinguishes different days of recording (can be `1` or `2` for patients with PD and is always `1` for patients without PD). All patients with PID <= 829 have Parkinson's Disease and all patients with PID >= 890 do NOT have Parkinson's Disease and hence belong to the control group. | |
| ### Fields in each File | |
| A `.mat` file can be read in python as follows: | |
| ```python | |
| from scipy.io import loadmat | |
| filename = "813_2_PD_REST.mat" | |
| mat = loadmat(filename, simplify_cells=True) | |
| ``` | |
| (A field "fieldname" can be read from `mat` as `mat["fieldname"]`.) | |
| Then `mat` contains (among others) the following fields and subfields | |
| - `EEG` | |
| - `data`: EEG data of shape `(#channels, time_len)`. E.g. a recording of 7 minutes with 67 channels and a sampling rate of 500 Hz will have shape `(67, 210000)`. | |
| - `event`: Contains a list of dictionaries, each entry (each event) having the following description: | |
| - `latency`: The onset of the event, measured as the index in the time-dimension `time_len`. The duration of each event is 1 second. Hence, with a 500 Hz sampling rate, the EEG data `event_data` corresponding to the `i`-th event would be | |
| ``` | |
| start_index = mat["EEG"]["event"][i]["latency"] | |
| event_data = numpy.transpose(mat["EEG"]["data"], [1, 2]).reshape([#channels, #trials * trial_len])[:, start_index:start_index+500] # shape (#channels, 500) | |
| ``` | |
| - `type`: The type of event. Can be `"S 1"` / `"S 2"` (patient has eyes open), or `"S 3"` / `"S 4"` (patient has eyes closed). | |
| - `chanlocs`: A list of channel descriptors | |
| - `nbchan`: Number of channels | |
| - `srate`: Sampling Rate (Hz) | |
| ## License | |
| By the original authors of this work, this work has been licensed under the PDDL v1.0 license (see LICENSE.txt). |