Functional near-infrared spectroscopy (fNIRS) is a brain imaging methodology that is appertaining to imaging the hemodynamic responses or blood oxygenation level-dependency of the brain. Machine learning (ML) has the potential to classify different mental states from the fNIRS signal to build a brain-computer interface (BCI) system. In this work, we have used a portable fNIRS system, WearLight, to image the prefrontal cortex of the brain of 12 participants while performing n-back working memory (WM) tasks. We converted the time-series raw data to the hemodynamic response by implementing a processing pipeline. A simple method based on checking the hemodynamic signal level of all the channels during task blocks with respect to the rest periods was used to identify the thee most dominant channels. We extracted eight important features from the hemodynamic brain signals to construct the feature matrix and trained six different k-nearest neighbors (k-NN) ML classifiers. The performance of the six k-NN classifiers was evaluated with the new experimental data sets. The results have shown that Weighted and Fine k-NN performed best (75%) in classifying the 4 different WM loads. The method is encouraging for real-time classification of the mental workload using a portable fNIRS system augmented by advanced machine learning techniques.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.