There are two predominant means of identifying visual areas in the human brain; retinotopy (exploiting maps of the visual field) and localisers (exploiting functional selectivity). This thesis aimed to bridge those two approaches, assessing the roles of LO-1 and LO-2; two retinotopically-defined regions that show overlap with the functionally-defined (shape selective) Lateral Occipital Complex (LOC). More generally, we asked what is the nature of the shape representation across Lateral Occipital cortex?
We first probed the functional roles of LO-1 and LO-2, finding that LO-2 is the more shape-sensitive region of the pair and will respond to second order shape stimuli, whereas LO-1 may process more local cues (perhaps orientation information).
Our later work then assessed neural shape representations across visual cortex, identifying two discrete representations; ‘Shape-profile’ (essentially retinotopic responses) and ‘Shape-complexity’ (responses based upon the complexity of a shape’s contour). The latter dimension captured variance in LOC, and surprisingly LO-2. This indicates that even explicit visual field maps can respond to non‑retinotopic attributes such as curvature complexity. Intriguingly, a transition between dimensions occurred around LO-1 and LO-2.
Finally, we explicitly tested whether the ‘Shape-complexity’ representation may be curvature based. Our results implied that radial shape protrusions are highly salient features for Lateral Occipital cortex, but it is not necessarily the points of maximal curvature that are being responded to. Instead, we hypothesise that it is the convergent lines comorbid with curvature that neurons may be attuned to, as such lines likely represent the most salient or characteristic features in a given shape.
In sum, we argue for an evolving shape representation across visual cortex, with some degree of shape sensitivity first emerging around LO-1 and LO-2. These maps may then be acting as preliminary processing stages for more selective shape tunings in LOC.
