Increasing the sensitivity of the multifocal visual evoked potential (mfVEP) technique: incorporating information from higher order kernels using a principal component analysis method.

Principal component analysis (PCA) was employed to enhance the multifocal visual evoked potential (mfVEP) technique. First, by using four principal components, the separation of mfVEP signals from noises was improved. Second, two otherwise unused higher order kernels, the 2nd slice of the 2nd order kernel and the 4th order kernels, were utilized by combining information obtained from the PCA. The PCA-kernel method improves the efficiency of the mfVEP test. The false positive rate, based upon an analysis of a noise window, was decreased by a factor of about one-third and the improvement in sensitivity of detecting glaucomatous defects was nearly as good as a doubling of the recording time.