Theses Doctoral

Neural Correlates of Early-Stage Visual Processing Differences in Developmental Dyslexia

Levinson, Lisa Merideth

Reading requires the successful recruitment and coordination of brain networks to translate visual symbols into phonemes, which are then sequenced to match speech sounds and matched onto semantic representations. Although phonemic awareness is understood to be a core deficit associated with reading disability, neuroimaging has demonstrated an association between poor reading and disruption to various interrelated areas in the brain. This includes one of the major visual pathways, the magnocellular pathway, which contributes to the dorsal pathway in the brain and the processing of motion. For at least two decades, researchers have observed differences in motion processing, supported by the magnocellular pathway, between individuals with and without dyslexia (Eden et al., 1996; Gori et al., 2016; Livingstone et al., 1991; Wilmer, 2004). Further, psychometric studies report an association between reading ability and dorsal stream sensitivity in adults and in children before and after learning to read (Boets et al., 2011; Kevan & Pammer, 2009). Studies of the development of the major visual pathways have suggested that the magnocellular pathway follows a protracted course of development, which raises the possibility that it is vulnerable to pathological change during development and also has the potential for greater plasticity (Armstrong et al., 2002; Stevens & Neville, 2006).
To explore the potential differences in early-stage visual processing, this dissertation study investigated whether neurophysiological measures, as indexed by event-related potentials (ERP), may differ between adults with and without dyslexia to stimuli tailored to evoke a response from each of two major visual pathways: magnocellular and parvocellular. The P1 component was elicited in response to motion stimuli designed to probe magnocellular pathways, and the N1 component was elicited in response to color stimuli designed for parvocellular processing. Group comparisons revealed statistically significant group differences in P1 amplitude for the motion/magnocellular condition, but no differences were found for N1 ERP measures for the parvocellular/color condition. Moderate to strong correlations between P1 measures in response to the magnocellular/motion condition were observed in relation to specific behavioral assessments: nonverbal reasoning and memory, orthographic choice, the word identification subtest from the Woodcock Reading Mastery Test (3rd edition: WRMT-III, Woodcock, 2011), and the sight word efficiency subtest from the Test of Word Reading Efficiency (2nd edition: TOWRE-2, Wagner, Torgesen, & Rashotte, 2011).
These results are indicative of an early-stage visual processing disruption in individuals with dyslexia observable at the level of the brain. Due to the compounding impact of even small disruptions of sensory and cognitive processing on learning, refining our knowledge of the underlying neural mechanisms of reading may permit earlier identification and potentially more focused interventions that could yield better outcomes for struggling readers. Additionally, the association of those differences with measures of word decoding will inform further research into the underlying neural mechanisms that may contribute to dyslexia and skilled reading.


  • thumnail for Levinson_columbia_0054D_14680.pdf Levinson_columbia_0054D_14680.pdf application/pdf 18.3 MB Download File

More About This Work

Academic Units
Speech and Language Pathology
Thesis Advisors
Froud, Karen
Ph.D., Columbia University
Published Here
May 16, 2018