fNIRS dataset of motor hand-gripping activity using the NIRSport2 system.

SIGNIFICANCE: The functional near-infrared spectroscopy (fNIRS) dataset acquired with the NIRSport2 device provides noninvasive recordings. The analysis of the fNIRS dataset can be used to refine existing models or propose new models to understand the motor cortex during voluntary motor activities, such as neuroplasticity and task-specific neural activation patterns. AIM: The hand-gripping dataset provides open-access fNIRS recordings of motor task-related brain activity to support the development and refinement of signal processing and machine learning models in brain-computer interface (BCI) research. APPROACH: Twenty healthy right-handed participants' fNIRS data is acquired during a hand-gripping task from the motor cortex using an 8 × 8 optodes configuration (twenty channels) following the international 10 / 20 system. The NIRSport2 device (NIRx Medizintechnik GmbH, Germany) is used to record hemodynamic cortical activity with a sampling frequency of 10.1725 Hz in the form of light intensity. Signal processing software nirsLAB (version: v201904_64bit) is used for preprocessing and converting the light intensity into optical density, which is then converted into hemoglobin concentration changes (oxy- and deoxyhemoglobin) and finally filtered out for physiological artifacts, consistent with our previous work on this dataset. RESULTS: This dataset is intended to explore patterns of brain activation during hand gripping, contributing to research on rehabilitation, motor learning, and neuroplasticity, and could be used to develop and validate classification algorithms, contributing to the field of BCI. CONCLUSIONS: These results demonstrate that the dataset is reliable and suitable for motor task analysis, benchmarking, and fNIRS-based machine learning studies.