Cortical activity measured with fNIRS related to freezing of gait in Parkinsons disease.

This dataset includes the raw data of 22 participants with Parkinson's disease and self-reported daily freezing, and 22 control participants performing a gait task including 180 degree turns, passing through a narrow doorway, stopping, and starting. One participant with Parkinson's did not give permission to share research data and is excluded from this dataset. Participants walked up and down a corridor, halfway passing through a narrow doorway frame (60 cm wide), and at the ends making 180° turns in a square taped on the floor (50 cm wide). At regular intervals, the participants were asked to stop and start walking again (see task_description.pdf). fNIRS data was sampled with a multichannel fNIRS device (Artinis Medical Systems, 48 long and 16 short separation channels) bilaterally over the primary motor cortex, the premotor cortex, the supplementary motor cortex, the prefrontal cortex, and the posterior parietal cortex (see optode_layout.pdf). The 3D positions of the optodes and anatomical landmarks were extracted from 3D scans of the participants' head (made with a Structure Sensor, Occipital). Motion data was collected with a full-body motion capture system (Xsens) consisting of 17 IMUs. The data of this dataset consists of the (.mvnx) exported data from the MVN Awinda software (version 2020.0.1) and includes acceleration, angular velocity and magnetometer data of each sensor ('sensor'), as well as positions, orientations, or other time derivatives of all body parts ('segments'). Additionally, center of mass position ('COM'), joint angles ('joint'), and foot contact data ('footcontact') are shared. The fNIRS and motion data was minimally processed (fNIRS data resampled from 50 to 60 Hz) and synchronized. Events were either extracted from the motion data, annotated based on videos, or extracted from the stimulus presentation file generated during the experiment (see events.json).