IC cells are BO cells!
DISC: a neuro-computational model
I developed a model for border-ownership (BOWN) computation, developed to reproduce illusory surface perception (DISC model, "differentiation and integration for surface completion" model, Kogo et al., 2010(1)). I hypothesized that the perception of illusory surface (Kanizsa square) is mediated by the emergence of BOWN signals at the gaps between the image elements (inducers) and that the emergent BOWN signals at the gaps constitute so-called illusory contour completion, or “modal” completion(2).
​
Hence, my hypothesis.
The illusory-contour-sensitive neurons, IC cells(3) are BOWN sensitive cells, BO cells(4).
​
I suggest that at first, the BO cells receive no sensory input at the gaps between the image elements. However, global computation at a higher level detects the configuration of an (illusory) surface. It then sends top-down signals to activate BO cells at the lower level, that are coherent with the existence of the surface. Consequently, the formerly silent BO cells at the gaps of the image elements now become activated. This emergent activation of silent BO cells corresponds to the perception of the illusory surface and its contours. It is possible that the interactions between the lower level and the higher-level show non-linear dynamics mediated by the NMDA receptor and dendritic activity.
Another type of completion phenomenon, “amodal completion”1, is perceived in the images used in the section for relatability in this paper. The output of the model indicates that BOWN signals of the border segments of the occluded surface at the intersection with the occluder are “grouped”, in the sense that they are considered to be consistent and enhance the interactions between them.
We take the position that, in the case of amodal completion, the above-mentioned activation of silent BO cells by feedback signals in modal completion does not occur due to the stoppage of the extension at the intersections with the occluder. This corresponds to the fact that we do not perceive modal completion of the contours at the location (i.e. we do not “see” the contours).
​
1. Kogo, N., Strecha, C., Van Gool, L., & Wagemans, J. Surface construction by a 2-D differentiation-integration process: A neurocomputational model for perceived border ownership, depth, and lightness in Kanizsa figures. Psychological Review 117(2), 406-439 (2010)
2. Michotte, A., Thinès, G. & Crabbé, G. Les compléments amodaux des structures perceptives (Amodal completion of perceptual structures.). (University of Leuven, 1964).
3. von der Heydt, R., Peterhans, E. & Baumgartner, G. Illusory contours and cortical neuron responses. Science 224, 1260–1262 (1984).
4. Zhou, H., Friedman, H. S. & von der Heydt, R. Coding of border ownership in monkey visual cortex. Journal of Neuroscience 20, 6594–6611 (2000).
​