Abstract
Purpose: :
Objective Perimetry for glaucoma diagnosis and early detection can be performed using pattern-reversal mfVECP. However, due to the complex structure and a high inter-individual variability, different parts of the visual cortex may contribute to the mfVECP signal in a very heterogeneous manner which is not easily predictable. In the present study, glaucoma detection was optimized by recording signals at several different locations on the skull to introduce different recording situations. Signal analysis was improved by a computer-based algorithm to find statistic parameters that allow for a distinction between normal and abnormal quadrants of the visual field.
Methods: :
mfVECP signal data was acquired with a four-channel measurement system (RetiScan Objective Perimetry, Roland Consult, Brandenburg, Germany) of 5 healthy volunteers and 5 patients with glaucomatous defects in different quadrants of the visual field. The visual Stimulus was an asymmetric dart board black/white reversal pattern with 56 segments (FOV 25°) and an additional lateral two-segment wing, extending the nasal FOV to 42°. A 19" LCD-TFT flat-screen monitor with a reaction time of 2 ms and a luminance of 430 cd/m2 was used as stimulator. Signal analysis was performed by a semi-automatic software tool based on customized MATLAB code (TheMathWorks, Natick, MA, USA) which allowed for the selection of the curves with the highest signal-to-noise ratio from the four recorded channels, digital signal-filtering and the computation of the P100-N135 amplitude differentials for each stimulated field and the quadrant and hemi-field mean values of the latter.
Results: :
By comparing quadrant- and hemi-field mean values, quadrants with glaucomatous defects could be identified (p < 0.05). A correlation between visual field loss in psycho-physic perimetry and the electrophysiological response pattern in the mfVECP was performed and resulted in a sensitivity of 0.82 and a specificity of 0.63 for quadrants with glaucoma-related visual field defects.
Conclusions: :
The optimization of signal post-processing in pattern-reversal mfVECP by computer-based automatic multi-channel curve selection, filtering and statistical group-forming yields a good detection of quadrants altered by glaucomatous visual field defects.
Keywords: electrophysiology: clinical • visual cortex • visual fields