Comparison of techniques for detecting visually evoked potential asymmetry in albinism

We compared techniques for analyzing visually evoked potential (VEP) asymmetry in children with albinism to find one that could be used effectively and efficiently.

Subjects included 21 child volunteers, ages 10 months to 6 years (control group) and 21 children with albinism, ages 2 months to 6 years (albinism group). Five-channel flash VEP was performed on all subjects. Electrodes were positioned at Oz, O1, O2, O3, and O4 (10/20 system). Data were analyzed by use of techniques previously described. These included inspection of the VEP waveforms, measurement of hemispheric waveform parameters, calculation of an asymmetry index, and use of a bipolar derivation between left and right hemispheric responses (interhemispheric difference potential). In addition, we quantified the interhemispheric difference potential by use of Pearson's correlation coefficient. Measurements of sensitivity and specificity determined the success of the 5 analysis paradigms. The accuracy of each paradigm represented the ability to classify the data according to volunteer or albinism group and is derived from both sensitivity and specificity measures.

Measurement of hemispheric differences in VEP waveform parameters was the least sensitive measure method for detecting multichannel VEP asymmetry in albinism. Comparison of left and right eye interhemispheric difference potential increased accuracy to 67%. Nonquantitative inspection of waveform demonstrated an accuracy of 76%. The asymmetry index and Pearson's correlate measure yielded accuracy rates of 79% and 83%, respectively.

The efficiency and capability of Pearson's correlate measure in quantifying interhemispheric difference potentials to detect albinotic misrouting makes this a useful and practical technique in a pediatric clinic.