Spatial summation across the visual field in strabismic and anisometropic amblyopia

www.nature.com/scientificreports OPEN Received: 25 September 2017 Accepted: 7 February 2018 Published: xx xx xxxx Spatial summation across the visual field in strabismic and anisometropic amblyopia Shindy Je1, Fergal A. Ennis1, J. Margaret Woodhouse1, Frank Sengpiel 2 & Tony Redmond 1 Ricco’s area (the largest area of visual space in which stimulus area and intensity are inversely proportional at threshold) has previously been hypothesised to be a result of centre/surround antagonism in retinal ganglion cell receptive fields, but recent evidence suggests a sizeable cortical contribution. Here, Ricco’s area was measured in amblyopia, a condition in which retinal receptive fields are normal, to better understand its physiological basis. Spatial summation functions were determined at 12 visual field locations in both eyes of 14 amblyopic adults and 15 normal-sighted controls. Ricco’s area was significantly larger in amblyopic eyes than in fellow non-amblyopic eyes. Compared to the size of Ricco’s area in control eyes, Ricco’s area measured significantly larger in amblyopic eyes. Additionally, Ricco’s area in the fellow, non-amblyopic eye of amblyopic participants measured significantly smaller than in control eyes. Compared to controls, Ricco’s area was larger in amblyopic eyes and smaller in fellow non-amblyopic eyes. Amblyopia type, binocularity, and interocular difference in visual acuity were significantly associated with inter-ocular differences in Ricco’s area in amblyopes. The physiological basis for Ricco’s area is unlikely to be confined to the retina, but more likely representative of spatial summation at multiple sites along the visual pathway. For a visual stimulus to be detected, the strength of the stimulus signal must overcome intrinsic noise that is inherent in the visual pathway. Pooling of signals over space (spatial summation) increases detectability, but at the expense of reduced visual resolution. Ricco’s law of spatial summation1 states that for a range of small stimulus areas, stimulus area (A) and intensity (I) are inversely proportional at threshold (A × I = k), i.e. spatial summation is complete. However, Ricco’s law applies only within a critical area, known as Ricco’s area. Beyond Ricco’s area, spatial summation is incomplete and, depending on the precise conditions under which it is measured, threshold is governed by laws of incomplete summation such as Piper’s law2 or Pieron’s law3. The physiological basis for Ricco’s area is not entirely understood. The traditional explanation has been that Ricco’s area reflects spatial antagonism in retinal receptive fields (as has similarly been hypothesised by Westheimer4 as the basis for the critical area in sensitization functions), but more specifically, that it is the psychophysical correlate of the area of the retinal ganglion cell (RGC) receptive field centre5,6. Wilson7 noted that spatial summation functions across the visual field could be superimposed by a simple displacement along the area axis, and that threshold for the largest stimulus undergoing complete spatial summation was invariant across the visual field. This was attributed to differences in RGC receptive field overlap across the visual field, based on the correlation between RGC density and receptive field centre size6,8. Initially, it may seem reasonable that Ricco’s area has a retinal basis, given that it has also been found to vary with retinal eccentricity7,9,10 and background adaptation level5,11 in healthy observers. However, despite the close association between Ricco’s area and RGC dendritic field size12, as well as eccentricity-related changes in RGC density9, Pan & Swanson demonstrated that spatial summation of circular incremental stimuli, as used in clinical visual field testing, cannot be accounted for by probability summation across retinal ganglion cells, but by cortical pooling by multiple spatial mechanisms13. Further support for the hypothesis that cortical pooling contributes to the physiological basis of Ricco’s area comes from Redmond et al., who found changes in Ricco’s area in the S-cone pathway as a function of blue background adaptation level14. The traditional explanation that changes in Ricco’s area with background luminance occurs due to increased spatial antagonism in RGC receptive fields5 cannot account for the results reported by Redmond et al.14 because centre-surround spatial antagonism is not found in receptive fields of the small 1 School of Optometry and Vision Sciences, Cardiff University, Cardiff, United Kingdom. 2School of Biosciences, Cardiff University, Cardiff, United Kingdom. Correspondence and requests for materials should be addressed to T.R. (email: ) Scientific REPOrtS | (2018) 8:3858 | DOI:10.1038/s41598-018-21620-6 1 www.nature.com/scientificreports/ bistratified cells that mediate S-cone signal response. Rather, the blue/yellow ON and OFF receptive field regions are spatially coextensive15. Receptive fields of the arrangement S+/S− would be required to observe such changes, and these are not found at the level of the retina. Additionally, compared to its size in age-similar healthy controls, Ricco’s area was found to be larger in patients with glaucoma14, a disease characterized by the death of RGCs. The traditional concept of Ricco’s area as strictly a retinal phenomenon fails to reconcile the apparent shrinkage of RGCs16,17 with the documented enlargement of Ricco’s area in glaucoma. Other contributions to Ricco’s area, such as cortical pooling, may explain this structure-function discordance in glaucoma. Since an enlarged Ricco’s area, such as occurs in glaucoma, can account for disproportionate deficits in contrast sensitivity to stimuli of different areas14, a better understanding of mechanisms other than retinal that contribute to Ricco’s area is essential to improve the design of functional visual field tests. The role of non-retinal contributions to Ricco’s area can be elucidated by determining whether a difference in Ricco’s area exists between eyes with normal vision and eyes with reduced vision in the absence of ocular or visual pathway pathology. Since amblyopia is a developmental disorder in which vision is reduced in the absence of detectable ocular or visual pathway disease, measuring spatial summation across the retina with stimuli of different areas in individuals with amblyopia may provide evidence for non-retinal contributions to Ricco’s area. Approximately 3.6% of the UK population has amblyopia18. Histological studies of experimentally-induced amblyopia have suggested that the primary site of developmental neural deficit is V119–22. RGCs have been observed to be anatomically and functionally normal (including normal spatial resolution23) in experimental models of amblyopia24–26. Although lateral geniculate nucleus (LGN) cells have been observed to change in size in severe deprivation amblyopia, their spatial resolution has been found to be unaffected19, (...truncated)