| Abstract: |
Assessing cognitive workload using functional near-infrared spectroscopy (fNIRS) in labs is well established. However, fNIRS sensors useful during normal activities in real-world environments are only recently emerging. We validated a small, portable fNIRS sensor (the fNIRS Pioneer ™) against a larger sensor with coverage of a larger cortical area, the NINScan developed at Massachusetts General Hospital. We used a gold-standard working memory task (n-back; (Kirchner, 1958)) and a more complex multi-attribute task battery (MATB) (Santiago-Espada et al., 2011). Twenty healthy adult (21.5 ± 3.3 years; 9 males) students at Brown University completed all three experimental visits. Fitting with previous research, on the n-back task, we found a significant effect of difficulty level on blood oxygenation (HbO2) in dorsolateral prefrontal cortex (dlPFC) HbO2 (p<.01), but not medial PFC HbO2 with the fNIRS Pioneer. For the NINScan, we observed increases in HbO2 from 1- to 2- to 3-back in two channels corresponding to the border between ventrolateral PFC (vlPFC) and dlPFC in both hemispheres (p<.05). When we aggregated MATB data across subtasks, and after accounting for time-on-task, we found a significant (p<.01) effect on HbO2 for the Pioneer and the NINScan. In all cases, the significant HbO2 findings were negative relationships, indicating less brain activation with better performance. While prior literature of functional brain imaging with MATB is not available, this finding is at least broadly consistent with the role of lateral PFC’s role in working memory. This indicates that both the fNIRS Pioneer and the NINScan sensor, when combined with appropriate data analytic techniques were useful for detecting changes in HbO2 that correlate with cognitive workload and behaviour, and that the fNIRS Pioneer is able to assess cognitive workload similarly to more larger, more expensive, and more established devices. |