Leveraging Functional Near-Infrared Spectroscopy for the Assessment and Prediction of Mental Workload

Evaluating workload is essential to maximizing performance with cognitively-aware interfaces. Neuroimaging modalities have been used to accomplish this task. The goal of the study is to leverage the efficacy of a neuroimaging technology which is safe, portable, affordable, and non-invasive to assess mental workload levels (MWL) and categorize as low and high. Functional Near-Infrared Spectroscopy technology is based on the measurement of hemodynamic responses obtained from the working memory-sensitive prefrontal cortex of the brain. It is considered to be a potential approach for a real-time window into cognitive demands. The study uses data from fifty people (25 Male and 25 Female) and the n-back task, which is a widely used paradigm for MWL analysis. An analysis has been conducted on the brain activation patterns related to changes in the concentrations of oxygenated (HbO) and deoxygenated (HbR) haemoglobin during different MWLs. Statistical features extracted from the brain activation patterns pertaining to different levels of workload are fed to classification model. A Graphical user interface (GUI) is developed as a means of feedback to the subject to showcase the level of workload as high or low. The developed GUI effectively predicts mental workload levels and displays corresponding images, allowing users to easily understand whether their workload is low or high based on visual feedback. This enhances the usability and accessibility of mental workload assessment.