Abstract
Goal Conflict Specific Rhythmicity (GCSR) is a biomarker of an anxiety process based on Gray’s neuropsychological theory of anxiety. It provides an objective definition of anxiety, empirical measure for diagnosis, and aims to align classification with cause. GCSR discriminates patients from non-patients at a group level but not at the individual level that is required for clinical use. This thesis aims to improve the individual discriminability of the biomarker. GCSR is measured via electroencephalography (EEG) in the stop signal task (SST). (Shadli, Glue, McIntosh, & McNaughton, 2015) showed that GCSR is not consistent across the SST. GCSR fluctuation rate (GCSR-FR) is a measure of GCSR variation across the task. GCSR-FR may be a better summary of GCSR than the current method. In this study there were four groups based on patient status and a theoretically driven manipulation effectiveness (ME) factor. The non-patient effective manipulation group (NP-EM) has 49 members, non-patient ineffective manipulation group (NP-IM) has 30, patient effective manipulation group (PEM) has 38, and the patient ineffective manipulation group (P-IM) has 19. GCSR-FR was extracted from the SST in three GCSR frequency bands, 4 – 6 Hz (theta), 7 – 9 Hz (low alpha), and 10 - 12 Hz (high alpha). The domain of GCSR-FR is 1 cycle per task to 16 cycles per task. The primary analysis revealed that patient factor differences in GCSR-FR in effective manipulation relative to patient factor differences in GCSR-FR in ineffective manipulation show a quartic (m or w shaped) pattern across the GCSR-FR domain in the theta and high alpha bands. The patient factor group differences in GCSR-FR that occur in effective manipulation groups but not in ineffective manipulation groups are good targets for refining patient status prediction and therein improving GCSR’s individual discriminability. We speculate that the quartic pattern occurs because of five unique trial strategies each preferentially aggregating in one of the four groups. The finding demonstrates that there is intra task variation in GCSR that traditional summaries are incapable of detecting. With further research, GCSR-FR can improve patient status prediction.