Assessment of cortical dysfunction in human strabismic amblyopia using magnetoencephalography (MEG)

The aim of this study was to use the technique of magnetoencephalography (M EG) to determine the effects of strabismic amblyopia on the processing of s patial information within the occipital cortex of humans. We recorded evoke d magnetic responses to the onset of a chromatic (red/green) sinusoidal gra ting of periodicity 0.5-4.0 c deg(-1) using a 19-channel SQUID-based neurom agnetometer. Evoked responses were recorded monocularly on six amblyopes an d six normally-sighted controls, the stimuli being positioned near the fove a in the lower right visual held of each observer. For comparison, the spat ial contrast sensitivity function (CSF) for the detection of chromatic grat ings was measured for one amblyope and one control using a two alternate fo rced-choice psychophysical procedure. We chose red/green sinusoids as our s timuli because they evoke strong magnetic responses from the occipital cort ex in adult humans (Fylan, Holliday, Singh, Anderson & Harding. (1997). Neu roimage, 6, 47-57). Magnetic field strength was plotted as a function of st imulus spatial frequency for each eye of each subject. Interocular differen ces were only evident within the amblyopic group: for stimuli of 1-2 c deg( -1), the evoked responses had significantly longer latencies and reduced am plitudes through the amblyopic eye (P < 0.05). Importantly, the extent of t he deficit was uncorrelated with either Snellen acuity or contrast sensitiv ity. Localization of the evoked responses was performed using a single equi valent current dipole model. Source localizations, for both normal and ambl yopic subjects, were consistent with neural activity at the occipital pole near the V1/V2 border. We conclude that MEG is sensitive to the deficit in cortical processing associated with human amblyopia, and can be used to mak e quantitative neurophysiological measurements. The nature of the cortical deficit is discussed. (C) 1999 Elsevier Science Ltd. All rights reserved.