**1. Introduction**

Dyslexia has been defined as a neurodevelopmental disorder characterized by reading difficulties in the absence of psychiatric, neurological, auditory or visual disabilities [1]. This disorder has been estimated to affect between 5 and 15% of children and around 4% of adults in the general population [2].

The pathophysiology of dyslexia is still controversial and has been attributed to phonological, auditory or visual alterations [3,4]. A number of neuroimaging investigations

**Citation:** Garcia-Medina, J.J.; Bascuñana-Mas, N.; Sobrado-Calvo, P.; Gomez-Molina, C.; Rubio-Velazquez, E.; De-Paco-Matallana, M.; Zanon-Moreno, V.; Pinazo-Duran, M.D.; Del-Rio-Vellosillo, M. Macular Anatomy Differs in Dyslexic Subjects. *J. Clin. Med.* **2023**, *12*, 2356. https://doi.org/10.3390/ jcm12062356

Academic Editors: Rodolfo Mastropasqua and Rossella D'Aloisio

Received: 11 February 2023 Revised: 13 March 2023 Accepted: 16 March 2023 Published: 17 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

focused on the central nervous system (CNS) have been performed so far to study normal [5] and abnormal reading process [2,6], and also reading interventions [7,8]. However, we have to keep in mind that the first steps of a successful reading process are entirely visual and are related to the retina, a part of the CNS located in the eye [9]. The retina is made up of a complex cell circuitry of neurons (photoreceptors, bipolar cells, horizontal cells, amacrine cells and ganglion cells) and glial cells (astrocytes, Müller cells and microglial cells) that are arranged into alternating layers of the nuclei and axons/synapses [9,10]. These layers can be individually segmented in vivo in cross-sectional scans, and their thickness quantified with a non-invasive and reproducible technology called optical coherence tomography (OCT), that achieves a high resolution in cross-sectional images (Figure 1a), similar to that obtained in histological sections [11]. OCT technology has been extensively used to study biomarkers in other CNS disorders [12,13].

**Figure 1.** Cross-sectional and en face scans obtained by the means of optical coherence tomography (OCT). (**a**) Automatic segmentation of the different intraretinal layers in a cross-sectional OCT image of the macula from a right eye. The fovea is depicted in yellow, the parafoveal subfield in purple and the perifoveal subfield in blue. See Figure 1b,c for the corresponding en face representation of the fovea, parafovea and perifovea with the same color code as the one used in this figure. Dashed green lines indicate the limits of the temporal perifovea, the temporal parafovea and the fovea in relation to the en face image (see also Figure 1b). Segmentations are also shown. RNFL = retinal nerve fiber layer, GCL = ganglion cell layer (GCL), IPL= inner plexiform layer, INL = inner nuclear layer, OPL = outer plexiform layer, ONL = outer nuclear layer, GCC = ganglion cell complex, MIDDLE = middle retinal layers, INNER = inner retina, OUTER = outer retina. (**b**) Early Treatment Diabetic Retinopathy Study (ETDRS) grid with concentric circles of 1, 3 and 6 mm diameters of the right eye, showing nine sectors of the macula in an en face OCT image. The OCT device automatically estimates the mean thickness in microns for each sector and for each segmentation (see also Figure 1a). T = temporal, N = nasal, S = superior, I = inferior, C0 = fovea. Number 1 and number 2 refer to the inner

ring and the outer ring, respectively, and correspond to the parafovea (inner ring) and the perifovea (outer ring). The macula is depicted with the same color code as in Figure 1a (the fovea in yellow, the parafovea in purple and the perifovea in blue). The horizontal, solid, green line indicates the location of the cross-sectional scan of Figure 1a in the en face image. Dashed green lines indicate the limits of the temporal perifovea, the temporal parafovea and the fovea in relation to cross-sectional images (see Figure 1a). (**c**) Early Treatment Diabetic Retinopathy Study (ETDRS) grid with concentric circles of 1, 3 and 6 mm diameters of the left eye, showing nine sectors of the macula in an en face OCT image. The OCT device automatically estimates the mean thickness in microns for each sector and for each segmentation (see also Figure 1a). T = temporal, N = nasal, S = superior, I = inferior, C0 = fovea. Number 1 and number 2 refer to the inner ring and the outer ring, respectively, and correspond to the parafovea (inner ring) and the perifovea (outer ring). The macula is depicted with the same color code as in Figure 1a (the fovea in yellow, the parafovea in purple and the perifovea in blue).

Specifically, the reading process starts with the projection of a well-focused image of the text onto the central part of the retina, called the macula. Then, this macular image is encoded by photoreceptors in a process called phototransduction, and transmitted through the chain of neurons of the visual pathway to the brain cortex to be interpreted [14]. The macular region includes the fovea, the parafovea and the perifovea [10] (Figure 1a–c).

Despite the macula being such an important site for the reading process, no study is available about the macular morphology in dyslexia as far as we know. Thus, the aim of this research was to compare the thickness of different retinal segmentations between dyslexic subjects and normal controls at the macula, by the means of OCT.

#### **2. Materials and Methods**

#### *2.1. Recruitment*

In this study, dyslexic and normal controls were prospectively recruited from the patients attending the hospital for a routine ophthalmic review at the General University Reina Sofia Hospital of Murcia, Spain.

The inclusion criteria for the dyslexic group were: (1) Caucasian race; (2) Spanish as mother tongue; (3) aged under 25 years; (4) previous diagnosis of dyslexia confirmed by at least two different specialists; (5) refraction less than 6 spherical diopters and 2.5 cylinder diopters; (6) visual acuity of 20/25 or higher; (7) no history or findings of eye diseases or previous eye surgery; (8) no extraocular disease capable of modifying OCT measurements; (9) a reliable OCT scan (see below); (10) no other sensory, neurological, psychiatric, emotional or intellectual disorders; (11) no socio/economic significant disadvantage. Self-reported normal reader controls had the same inclusion criteria, except for criterion number 4. Only those participants not self-reporting reading difficulty and who correctly read aloud a simple 5-line paragraph text, without making a mistake or awkward pauses, were recruited in the control group.

#### *2.2. Ophthalmic Examinations*

All the patients underwent a complete ophthalmic examination in both eyes, including visual acuity, autorefraction, air pneumatic tonometry, biomicroscopy and funduscopy. If the candidates were eligible, they underwent posterior pole horizontal protocol with 768 Aand 61 B-scans taken 30 × 25 degrees centered at the fovea, using a Spectralis OCT spectral domain device with an eye-tracking system (software version 6.0; Heidelberg Engineering, Heidelberg, Germany).

The OCT examinations were performed in the morning, between 8:30 and 12:30 h, with pupil dilation. During this examination, the mean thickness of nine sectors was determined for each considered segmentation using the 1, 3, 6 mm diameter Early Treatment Diabetic Retinopathy Study protocol, that is one of the most used OCT en face patterns to study the macula, which considers the fovea, parafovea and perifovea subfields (Figure 1b,c). The position of the fovea was automatically determined by the device and checked by the same ophthalmologist (J.J.G.M.). Only reliable scans, with a signal strength over 20, were

included. All the scans were performed by the same operator and inspected by the same experienced ophthalmologist (J.J.G.M.), in order to exclude eyes with segmentation errors, decentering or any other artifact. No manual corrections were made to the automatic segmentation performed by the prototype software. The examinations with decentrations, or with segmentation errors that could alter thickness estimation at any sector, were excluded.

Then, with a segmentation tool (Segmentation Technology; Heidelberg Engineering), the thickness values of the following segmentations were automatically obtained: complete retina, inner retina, outer retina, retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL). The thickness value of the other segmentations was also obtained by summing up the thicknesses of the automatically obtained segmentations as follows: ganglion cell complex (RNFL + GCL + IPL), middle retina (INL + OPL + ONL), IPL + INL and OPL + ONL (Figure 1a). The thicknesses of OPL and ONL were not considered individually and were summed up, as was performed in previous works regarding autism [15] and glaucoma [16–18], because these two layers are hard to be conveniently separated in OCT images due to their similar reflectivity. A three-dimensional video (Video S1) has been made in order to better understand the integration of cross-sectional (Figure 1a) and en face OCT images (Figure 1b,c). Considering that neurons and glia in the human retina are organized in concentric rings around the fovea [19,20], we also summarized the thickness for all the mentioned segmentations considering the inner ring subfield ((S1 + N1 + I1 + T1)/4) and the outer ring subfield ((S2 + N2 + I2 + T2)/4) that correspond to the parafovea (purple ring in Figure 1b,c) and to the perifovea (blue ring in Figure 1b,c), respectively.
