A Temporal Sampling Basis for Visual Processing in Developmental Dyslexia

Abstract

Knowledge of oscillatory entrainment and its fundamental role in cognitive and behavioral processing has increasingly been applied to research in the field of reading and developmental dyslexia. Growing evidence indicates that oscillatory entrainment to theta frequency spoken language in the auditory domain, along with cross-frequency theta-gamma coupling, support phonological processing (i.e., cognitive encoding of linguistic knowledge gathered from speech) which is required for reading. This theory is called the temporal sampling framework (TSF) and can extend to developmental dyslexia, such that inadequate temporal sampling of speech-sounds in people with dyslexia results in poor theta oscillatory entrainment in the auditory domain, and thus a phonological processing deficit which hinders reading ability. We suggest that inadequate theta oscillations in the visual domain might account for the many magno-dorsal processing, oculomotor control and visual deficits seen in developmental dyslexia. We propose two possible models of a magno-dorsal visual correlate to the auditory TSF: (1) A direct correlate that involves "bottom-up" magnocellular oscillatory entrainment of the visual domain that occurs when magnocellular populations phase lock to theta frequency fixations during reading and (2) an inverse correlate whereby attending to text triggers "top-down" low gamma signals from higher-order visual processing areas, thereby organizing magnocellular populations to synchronize to a theta frequency to drive the temporal control of oculomotor movements and capturing of letter images at a higher frequency.

Keywords: dorsal; dyslexia; gamma; magnocellular; oscillations; reading; temporal sampling; theta.

FIGURE 1

Comparison of eye-movements during reading. (Left image): Normal reader. Fixations (represented by dots) and saccades (represented by the connecting lines) occur in a linear fashion at an approximate temporal frequency of 4 Hz. (Right image): Dyslexic reader. Fixations (represented by dots) occur erratically with no rhythmic temporal pattern (figure adapted from Prado et al., 2007).

FIGURE 2

Stylized representation of Goswami’s (2011) auditory temporal sampling framework (TSF). Syllabic sounds in spoken language occur at a theta frequency. It is hypothesized that when neuronal populations in the auditory domain attend to speech-sounds they phase-lock (entrain) to this theta frequency. Figure adapted from Pammer (2014).

FIGURE 3

Stylized representation of a possible visual correlate to the auditory temporal sampling framework. It is hypothesized that in normal reading theta frequency fixations entrain theta brainwave activity in the visual domain, thus supporting cognitive visual processing during reading (figure adapted from Pammer, 2014).

FIGURE 4

Summary of the rationales for theta-frequency eye-movements during reading, which may be orchestrated by low-frequency claustral oscillatory output.

FIGURE 5

Representation of a hypothesized bottom-up model of visual temporal sampling during normal reading (Model 1). (Top figure): Normal reading. 1. Eye-movements during reading occur at a theta-frequency and act as an entrainment stimulus. 2. Magnocellular oscillations in visual sensory areas phase-lock to the theta-rhythm of eye-movements, thus enabling visual coding of text. 3. Theta phase-locked oscillations drive cross-frequency oscillatory coupling to gamma. 4. Gamma oscillations become nested within entrained theta oscillations, thus enabling transfer of information to higher-order visual areas for processing of text. (Bottom figure): Reader with dyslexia. 1. Dyslexic readers have erratic eye-movements during reading, generating no steady rhythm of visual stimulus shifts. 2. The action potentials of magnocells respond to each stimulus shift. 3. However, a lack of theta eye-movements means there is no stable rhythm to which magnocellular populations can entrain and this impairs coding of text. 4. Lack of theta synchronization in visual sensory areas means that cross-frequency coupling to gamma cannot occur, thus hindering communication of information along the magno-dorsal pathway 5. Lack of increased gamma synchronization in the PPC results in impaired processing of text.

FIGURE 6

The magno-dorsal circuit. During bottom-up visual processing during reading signals from retinal magnocells are projected to magnocellular layers of the lateral geniculate nucleus (LGN). From the LGN, signals are projected from the LGN to V1 (visual cortex), where, the visual pathway diverges into the dorsal (“where”) and ventral (“what”) streams. The dorsal pathway, dominated by magnocells, is constituted by a hierarchy of cortical areas, namely, V2, V3, MT/V5 and the posterior parietal cortex (PPC). In top-down visual attention during reading, neuronal signals are sent from the PPC to MT/V5, V2, V1 and LGN. The PPC and MT/V5 also project to the pre-frontal cortex (PFC), which plays a key role in the control of eye-movements during reading. The ventral stream, proceeds to V4 and further on to the inferior temporal cortex (ITC).

FIGURE 7

Representation of the hypothesized magno-dorsal temporal sampling framework as a top-down and bottom-up process (Model 2). (Top figure) Normal reading: When visual attention in the posterior parietal cortex (PPC) is directed to text it drives a top-down modulation of oscillatory activity. Oscillations in the PPC synchronize to gamma and drive high-to-low cross-frequency coupling to theta. When oscillations in visual sensory areas synchronize at a theta frequency it causes eye-movements during reading to also occur at a theta frequency. (Bottom figure) Reader with dyslexia: When visual attention is directed to text, oscillations in the PPC fail to synchronize effectively at a gamma frequency. This in turn hinders also the effectiveness of the PPC in modulating top-down signals required for theta-controlled eye-movements during reading. Erratic eye-movements mean visual areas lack theta-phase-locking reinforcement, thus interrupting consolidation of theta-gamma activity and hindering bottom-up information transfer.