The human visual system is far more complex than a single cable from eye to cortex. Visual signals travel from the retina through two major routes: the retinogeniculate system, which passes through the lateral geniculate nucleus of the thalamus and divides into magnocellular, parvocellular and koniocellular pathways, and the retinotectal system, which projects to the superior colliculus and then to the inferior pulvinar nucleus. Each pathway is tuned to different stimulus properties. Magnocellular neurons respond most strongly to achromatic, low-spatial-frequency information and support motion and rapid temporal processing. Parvocellular neurons are uniquely sensitive to color contrast and fine spatial detail. Koniocellular cells, activated preferentially by short-wavelength blue-yellow opponency, innervate only the superficial layers of primary visual cortex, where they act as modulators rather than drivers of cortical activity. Approximately ten percent of magnocellular and koniocellular retinal ganglion cells bypass the lateral geniculate entirely, targeting the superior colliculus and inferior pulvinar through the retinotectal route.

Decades of research have established that people with schizophrenia show preferential deficits in processing stimuli that bias the magnocellular pathway, and the most parsimonious explanation involves impaired N-methyl-D-aspartate receptor (NMDAR) function within subcortical and early visual regions. Recent post-mortem work reinforces this idea, demonstrating substantial reductions of NMDAR expression specifically within primary and secondary visual cortical areas in schizophrenia, alongside a caudal-rostral gradient of histological disturbance that is more extensive in visual cortex than in anterior brain regions. Yet the specialized equipment needed for many of these assessments — contrast sensitivity paradigms, visual backward masking, pattern-reversal VEP — cannot easily be deployed outside dedicated research laboratories, leaving a critical gap between what science knows and what clinical practice can measure.

The new study, led by Maria B. Aburto-Ponce, Kristin Micceri, Antigona Martinez and Daniel C. Javitt, set out to close that gap by exploiting flash-VEP, a procedure routinely obtained with standard clinical EEG systems, most commonly to assess epileptogenic potential. The researchers recorded flash-VEP from 28 healthy controls and 27 people with schizophrenia in response to white and red photic stimuli delivered by an LED-based device at stimulation rates of 1, 3, 6, 9 and 12 Hz, and pattern-VEP from 29 controls and 22 schizophrenia participants in a parallel feature-attention study, with 17 controls and 15 patients participating in both. An additional cohort of 14 controls and 14 patients underwent fMRI to 6-Hz white and red flicker, with analyses focused on dorsal and ventral visual cortical parcels and the inferior pulvinar nucleus. Crucially, the team applied contemporary time-frequency decomposition, using a three-cycle Morlet wavelet analysis across 73 logarithmically spaced frequencies, to separate evoked power, total power, induced power and intertrial coherence (ITC), also known as phase locking.

The results revealed a striking double dissociation. People with schizophrenia showed significantly increased flash-tVEP responses, with a large effect size of d = 0.99 (p < 0.001), yet significantly reduced pattern-tVEP responses in the same direction of magnitude, d = −0.99 (p < 0.001). A mixed model regression across both stimulus types confirmed a highly significant group-by-stimulus-type interaction (F(1,102) = 23.9, p < 0.001, d = 0.97), and within the subset who completed both studies the interaction was even larger (d = 1.9). This dissociation is physiologically meaningful. In non-human primates, the flash-tVEP depends primarily on parvocellular inputs to visual cortex, because local inhibitory feedback within the superficial layers of primary visual cortex selectively suppresses magnocellular input; when that inhibition is blocked with bicuculline, a large localized increase appears in magnocellular-recipient layers, and that increase is in turn inhibited by the NMDAR antagonist ketamine. Red stimuli further enhance the parvocellular bias by preferentially activating the ventral visual stream. The enhanced flash-tVEP in schizophrenia, especially to red stimuli, may therefore serve as a sensitive index of pathological cortical remapping, including aberrant parvocellular input into dorsal stream pathways that normally receive low-spatial-frequency magnocellular drive.

Equally important were the deficits in the steady-state, or photic driving, response. When stimuli are delivered at rates of roughly 3 per second or faster, the primary EEG response consists of entrainment, or phase-reset, of ongoing posterior alpha rhythms, with no accompanying change in induced power or cortical activation. The researchers quantified this response by extracting the alpha-band ITC, converting it into a continuous pseudo-ERP variable and then performing a Fast Fourier Transform on the extracted response. For white stimuli, this analysis revealed a highly significant main effect of diagnostic group (F(1,53) = 7.67, p = 0.008, d = 0.78), with the greatest reduction observed at 6-Hz stimulation (d = 0.76). Notably, no between-group differences emerged for red stimuli, and there were no significant differences in alpha power across colors or stimulation rates — only in the coherence, or phase locking, of the response. The authors argue that this differential sensitivity to white versus red stimuli, together with the absence of a power change, is consistent with koniocellular involvement, given the insensitivity of the koniocellular system to long-wavelength red stimuli and its innervation only of superficial cortical layers. If the koniocellular system does mediate the photic driving response, this study would constitute the first neurophysiological demonstration of koniocellular impairment in schizophrenia, although one prior study reported reduced contrast sensitivity to konio-biased stimuli. Ketamine has been shown to suppress the ssVEP response in non-human primates, suggesting NMDAR involvement in koniocellular as well as magnocellular pathway function.

Perhaps the most historically resonant finding concerns alpha blocking, the sustained suppression of the posterior alpha rhythm that follows a flash stimulus. This phenomenon was first reported almost a century ago, shortly after the initial description of the EEG itself, and deficits in alpha blocking were documented in schizophrenia as early as 1938 and replicated repeatedly in the classic literature. Yet those early studies relied on qualitative approaches in which researchers manually counted alpha waves before and after stimulation. The new work represents the first quantitative study of alpha blocking in schizophrenia using modern analytic methods. Pre-stimulation amplitudes did not differ between groups, but activity in the 5–10 Hz range during the first five seconds of stimulation was significantly larger in patients (F(1,53) = 9.53, p = 0.003, d = 0.84), indicating that the normal flash-induced suppression of alpha was substantially reduced. In healthy controls, blocking was most prominent within the 6–10 Hz range over ventral cortex and was induced equivalently by white and red stimuli, suggesting magnocellular involvement via the retinotectal system; in schizophrenia, blocking was virtually absent.

The clinical significance of these electrophysiological deficits became clear when the researchers correlated them with neurocognitive performance measured using the MATRICS Consensus Cognitive Battery. Composite T-scores were dramatically reduced in schizophrenia (23.2 ± 12.0) relative to controls (44.1 ± 11.2), and the degree of alpha blocking correlated significantly with the composite score across groups (rp = 0.39, p = 0.007), with no significant difference in the slope of the relationship between diagnostic groups. Independent correlations emerged for Speed of Processing (rp = 0.49, p < 0.001), Visual Learning (rp = 0.47, p < 0.001), Verbal Learning (rp = 0.54, p < 0.001) and Working Memory (rp = 0.34, p = 0.021). Within Speed of Processing, the correlation was significant for the visual Symbol-Coding component (rp = 0.65, p < 0.001) but not for the non-visual Category Fluency test (rp = 0.18, p = 0.22), and within Working Memory it was significant for Spatial Span but not Letter-Number Span. The authors suggest that stimulus-induced alpha blocking in the EEG may specifically index the failure to engage visual systems during visually based tasks in schizophrenia, positioning dynamic alpha modulation as a key neurophysiological process related to cognitive processing.

The fMRI component provided converging anatomical evidence. In controls, activation was significantly greater in ventral than dorsal cortical regions (F(1,13) = 7.60, p = 0.016, d = 1.53), but this difference was absent in schizophrenia, reflecting a significantly larger response in patients to white stimuli within the dorsal visual stream (F(1,26) = 6.21, p = 0.02, d = 0.98) — an aberrant hypersensitivity consistent with the team’s earlier reports of functional reorganization in which dorsal stream cortex receives unexpected high-spatial-frequency parvocellular input. Within the pulvinar nucleus, controls showed significant activation to both white (d = 0.66) and red (d = 0.58) stimuli, whereas patients did not (both p > 0.20), yielding a highly significant main effect of group (F(1,26) = 8.14, p = 0.008, d = 1.12). When analyzed across regions and colors, the group-by-region effect was significant (F(2,25) = 7.04, p = 0.003, d = 1.5), capturing the combination of enhanced cortical and reduced subcortical activation in schizophrenia. The pulvinar deficit was equivalent for white and red stimuli, mirroring the color-independence of the alpha blocking deficit and supporting a potential role for magnocellular projections from the lateral geniculate to the inferior pulvinar in mediating the blocking abnormalities. The pulvinar is considered a key structure for alpha-frequency interactions between thalamus and cortex, capable of mediating both bottom-up and top-down modulations of ongoing alpha activity.

The authors acknowledge limitations, including an approximate rather than exact luminance match between the red and white conditions — the filters permitted 20 percent transmission for red-only light versus 10 percent for cross-colors — so differences could partly reflect general achromatic versus chromatic processing rather than specific pathway selectivity, although retinotectal afferents are known to be composed primarily of magnocellular and koniocellular projections, limiting the impact on retinotectal function. Flicker photometry could provide more accurate differentiation among the M, P and K pathways, but requires specialized equipment that is inherently less scalable to clinical settings. Control analyses showed no significant correlations between the ERP or fMRI measures and age, sex, medication dose or length of illness, and the one nominally significant correlation with sex did not alter the between-group difference. Taken together, the study proposes that the photic driving and alpha blocking responses to repetitive flash stimulation represent robust, computationally tractable signatures of early visual processing dysfunction in schizophrenia that correlate with overall levels of cognitive impairment and support concepts of subcortical contributions to cortical-level dysfunction. Because flash-VEP can be obtained with off-the-shelf clinical-grade EEG systems, the researchers argue that these visual biomarkers can now be translated from research laboratories into clinical settings, providing a scalable mechanism for assessing early-visual dysfunction and a concrete target for treatment development aimed at the NMDAR-mediated gain control that appears to lie at the heart of the illness.

Subject of Research: Subcortical visual pathway dysfunction and its relationship to cognition in schizophrenia, assessed with flash and pattern visual evoked potentials and fMRI of the pulvinar nucleus

Article Title: Differential deficits in pattern- versus flash-visual evoked potentials in schizophrenia: relationship to subcortical visual systems, pulvinar nucleus and cognition

Article References: Differential deficits in pattern- versus flash-visual evoked potentials in schizophrenia: relationship to subcortical visual systems, pulvinar nucleus and cognition. (n.d.). https://doi.org/10.3389/fnimg.2026.1833352

Image Credits: AI Generated

DOI: 10.3389/fnimg.2026.1833352

Keywords: schizophrenia, visual evoked potentials, magnocellular pathway, koniocellular pathway, pulvinar nucleus, alpha blocking, photic driving, NMDA receptors, EEG, fMRI, neurocognition, biomarkers