A Functional-Near Infrared Spectroscopy Investigation of Mental Workload
MetadataShow full item record
Mental workload assessment is a critical aspect of human-computer interfaces. Mental workload which is either underwhelming or overwhelming will result in a negative influence on one’s performance. Enhancing existing classification techniques of mental workload holds great potential in improving our understanding of cognitive engagement. Accurate detection of mental workload can have wide-ranging applications such as improving cognitive engagement or cognitive workload during an online class, which in turn affects the student’s performance and learning. In this research, we utilize functional near-infrared spectroscopy (fNIRS) to obtain insights into the dynamic functional connectivity of the brain as a function of the mental workload. Variations in the brain's blood oxygenation and deoxygenation reflect neuronal activation patterns and can be measured using fNIRS. Interpreting connectivity in the brain using noisy fNIRS data with low signal to noise ratio is challenging. To overcome the challenges with fNIRS data, we use a hierarchical latent dictionary learning approach. This approach provides covariance matrices to obtain the dynamic functional connectivity and neuronal activation patterns that change over time. We use features from the dynamic functional connectivity of the brain reflected in fNIRS data collected from the prefrontal cortex to investigate mental workload. An analysis of two different datasets, which use memory task consisting of different levels of difficulty, is presented. Covariance matrices are obtained for each difficulty level of the memory-based task by modeling the underlying neural signals as Gaussian processes and applying a latent factor model to the observed hemodynamic data, which in turn reflects the dynamic functional connectivity of the brain. The obtained covariance matrices are inputted to three machine learning algorithms, namely, Support Vector Machines, K-Nearest Neighbors, and Linear Discriminant Analysis, in order to evaluate classification accuracies in discerning the levels of difficulty. This hierarchical latent dictionary learning approach is implemented on an open-access dataset as well as a novel memory-related mental workload experiment conducted as part of this research. To elucidate the effects of music on mental workload, the novel experiment included calming and vexing music sessions. Our classification accuracies verify the viability of hierarchical latent dictionary learning approach to obtain functional connectivity and also expound the effects of music on mental workload.