Diagnostic value of automatically interpreted oculopneumoplethysmography

The diagnostic value of the Gee-oculopneumoplethysmography test for the detection of hemodynamically significant carotid artery obstructions has frequently been questioned due to the rather low agreement with arteriography. The kappa-corrected diagnostic accuracy of this test has been reported to be in the range between 0.54 and 0.68. These values indicate a "fair to good" agreement with arteriography. One of the causes for this rather low agreement is the variability in the interpretation of oculopneumoplethysmography recordings. Automatic analysis of oculopneumoplethysmography recordings might minimize the risks of misinterpretation and might improve the clinical significance of the Gee-oculopneumoplethysmography test. By assessing the signals of a conventional Gee-oculopneumoplethysmography apparatus in a personal computer, this automatic analysis has been realized. In a clinical study (population of 85 patients) the diagnostic value of automatic analysis of oculopneumoplethysmography recordings has been evaluated. The automatically interpreted Gee-oculopneumoplethysmography test had a kappa-corrected diagnostic accuracy of 0.71 on a "per side" base and of 0.76 on a "per patient" base.