Assessing concussion with multifocal pupillographic objective perimetry

Purpose: Recent results indicate that multifocal pupillographic objective perimetry (mfPOP) drives the cortical input to the pupillary system. Here we compare diagnostic power of 3 mfPOP variants in athletes who have had concussive injuries. Methods: The 3 dichoptic mfPOP variants included a version with 44 test regions/eye that had a test duration of 6 minutes, and two 80-second tests with 12 test regions/eye. The two 80-second tests had maximum luminances of 156 and 216 cd/m2. We tested male subjects: 18 control subjects aged 24.3 ± 4.09 y (mean ± SD), and 37 rugby athletes with prior concussion aged 21.6 ± 2.11 y. We categorised the concussion patients into tertiles according to: days since concussion, concussions in the last year, and life-time number of concussions. All the concussions caused removal from the playing field. We conducted ROC analysis on the using regional sensitivities, delays and linear discriminant scores combining sensitivities and delays. Results: The best discrimination was obtained between controls and persons who had concussion between 4 and 34 days before testing for the dimmer 80-second test, using the combined scores resulting in an area under the ROC curve of 0.85 ± 0.07 (mean ± SE). Performance was about equal for the mean of the worst 4 to 6 regions/eye. Thus the best discrimination was for a small number of defective visual field regions. Conclusions: Reasonable diagnostic power was obtained for recently concussed athletes. Studies with more subjects, and more recent concussion, are needed. Nevertheless the results were encouraging.