Objective detection of hemifield and quadrantic field defects by visual evoked cortical potentials

An objective method for detecting hemifield and quadrantic visual field defects has been developed using steady state visual evoked cortical potentials (VECPs), an adaptive noise canceller (ANC), and Hotelling's t2 statistic. The purpose of this study was to determine the sensitivity and specificity of the technique.

Nine subjects (mean age 44 years) were investigated with field loss due to a variety of causes including both anterior and posterior visual pathway lesions. Dynamic perimetry was performed by means of a Goldmann or Tübingen perimeter. VECP recordings were made from each visual field quadrant (23 degrees X 23 degrees) by means of a steady state reversing checkerboard (7.7 rev/s). The central 5 degrees of the visual field and the vertical and horizontal meridians were masked during these measurements. Recordings were made from three electrode sites, positioned over the visual cortex, relative to a mid frontal electrode. Each recording lasted 2 minutes, during which time fixation was monitored. The data from each recording were divided into 4 second segments, and the amplitude and phase of the VECP signal measured using the ANC. Hotelling's t2 statistic was applied to determine the probability of signal detection. Receiver operating characteristic curves were used to find the optimum signal detection threshold for identification of the visual field defects.

The results of the study confirmed patterns of subjective visual field loss. The technique had a sensitivity and a specificity of 81% and 85%, respectively, for detecting 'non-seeing' areas in the inferior visual field, and 82% and 89%, respectively, for detecting 'non-seeing' areas in the superior visual field.

These results demonstrate that the technique is of potential clinical value to ophthalmologists and neurologists when subjective perimetry is not possible.