Neuroergonomics and Vigilance: Probing the Event-Related Cerebral Hemodynamic Response

Abstract

Vigilance is an operator’s ability to sustain attentional focus and remain alert for a prolonged period of time in order to monitor for and detect critical task-relevant stimuli (Davies & Parasuraman, 1982; Warm, Finomore, Vidulich, & Funke, 2015). The primary subject of vigilance research is the vigilance decrement, a temporal decline in performance that characterizes vigilance tasks. The vigilance decrement poses a threat to safety and human effectiveness in operational settings where vigilance is required, of which there are many, including aviation, healthcare, and military domains (Davies & Parasuraman, 1982). Consequently, ongoing research efforts emphasize the need to gain a better understanding of the vigilance decrement and to develop countermeasures. One compelling explanation for the vigilance decrement is resource theory, which proposes that the information processing demands of vigilance tasks are so demanding that they deplete limited neurocognitive resources necessary for vigilance at a rate that is outpaces resource replenishment (Davies & Parasuraman, 1982). As the resources necessary for vigilance declines, so too does vigilance performance. Support for resource theory has been derived from a wealth of behavioral research showing that as the demands of a vigilance task are increased, performance tends to decline more rapidly (for review, see Warm et al., 2015). Indeed, more difficult vigilance tasks appear to increase the rate at which resources are depleted, which increases the rate at which performance is impaired. Further evidence in favor of resource theory comes from neuroergonomic studies that facilitate more direct measurement of neural resource utilization by employing neurophysiological measurements. Transcranial Doppler sonography (TCD) is one such measure. TCD provides a measure of cerebral blood flow velocity (CBFV), which is a metabolic index of cortical activity and neural resource utilization (Toole, 1984; Tripp & Warm, 2007; Warm, Tripp, Matthews & Helton, 2012). Vigilance research using the TCD has shown that CBFV tends to decline in parallel with vigilance performance (Warm & Parasuraman, 2007; Warm et al. 2012). Further, CBFV is greater, initially, when vigilance task demands are increased, but increases in task demands also lead to a more precipitous decline in CBFV – effects that align with a resource theory account (Warm & Parasuraman, 2007; Warm et al. 2012). Those findings, like most research using the TCD in vigilance, involved examination of CBFV that has been averaged across large (e.g., 10-min) periods of time. A recently developed alternative to that global method of CBFV analysis, is event-related CBFV analysis, which involves a temporally nuanced examination of the CBFV response in the seconds following certain types of stimulus presentations and performance outcomes (e.g., correct detections, correct rejections). A comparatively small amount of research has been done on event-related changes in cerebral blood flow velocity. The initial investigation was conducted by Shaw and colleagues (2013) who demonstrated that CBFV increases immediately after an individual detects a critical signal and that this CBFV increase is absent when an individual correctly rejects a neutral (non-target) stimulus. They also found that the magnitude of that detection-specific response declined over time, similar to global CBFV and the vigilance decrement. These findings led Shaw and colleagues (2013) to suggest that event-related CBFV is sensitive to the activity and depletion of detection-specific resources that are involved in correctly detecting a critical signal. In a related study, Greenlee and colleagues (2016) replicated the detection-specific CBFV response using event-related analyses. They also showed that event-related CBFV may be more sensitive than global CBFV to the demands of vigilance tasks. The current study was conducted to provide further examination of that possibility. Specifically, we aimed to assess the sensitivity of event-related CBFV and global CBFV to differences in signal probability, a well-studied variable known to influence attentional demands in vigilance tasks (Warm et al., 2015). We expected resource demands to be greater in a low signal probability condition than in a high signal probability condition. Consequently, we predicted that the magnitude of the detection-specific response, the magnitude of global CBFV, and the severity of the decline in both measures would be greatest in the relatively demanding, low signal probability condition. Twenty-four undergraduate students (22 Women, 2 Men) were assigned at random to either a low critical signal probability condition (3.3%) or a high critical signal probability condition (6.6%). All participants completed a 40-minute computerized vigilance task adapted from Greenlee and colleagues (2015). Participants were tasked with monitoring a computerized gauge which was updated 30 times per minute, and they were instructed to respond when the gauge needle deviated from a vertical position (neutral signal) to a rotated position (critical signal). Analysis of performance data indicated that correct detection rates declined significantly over the course of the vigil, but there was no performance difference between high and low signal probability conditions. Global CBFV mirrored these performance data with a significant decline in global CBFV, but no difference between conditions. Analyses of event-event related CBFV revealed that CBFV increased in the seconds following a critical signal that was correctly detected, but not after a neutral signal was correctly rejected. While this represents a detection-specific response that was similar to previous reports (Greenlee et al., 2016; Shaw et al., 2013), the magnitude of that response did not decline over time and did not vary between the two conditions. In sum, the present results do indicate that event-related and global CBFV may be differentially sensitive to the demands of vigilance, but not in the manner reported previously. Given that global CBFV and performance declined in a similar fashion and event-related CBFV did not decline, it appears that global CBFV may have been more sensitive to resource utilization and depletion within the current vigilance task. This finding contrasts with those of Greenlee and colleagues (2016), who found evidence that event-related analyses may be more sensitive. More research will be required to determine the source of this discrepancy between studies and to more completely evaluate the relative value of global and event-related analyses of CBFV during vigilance tasks.