Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closewhat happens around the fovea?
Posted by sydneyunivisphys on 08 Jun 2007 at 23:19 GMT
The model requires an offset in the spatial position of ON- and OFF-arrays. In high-acuity, diurnal primates such as macaques, the centre of ON- and OFF cells in the midget-parvocellular pathway gets input from a single cone (and one cone gives output to both ON- and OFF pathways), up till at least 8 degrees from the fovea, presumably meaning the offset is not present in these arrays. Further, the virtual absence of ganglion cells from the foveal pit causes a strong distortion in the arrays. Orientation maps are nevertheless present in primate V1 in the parafoveal region. A prediction of this model therefore seems to be that in primate V1, parvocellular LGN inputs to do not contribute to the formation of orientation maps, and there should be little or no orientation map in the parvocellular input layers. Apart from Poggio’s 1972 paper (Invest Ophthalmol 11 (5) 368-377) finding somewhat fewer orientation selective-cells near fovea is there strong evidence for or against this?
RE: what happens around the fovea?
dario replied to sydneyunivisphys on 24 Jul 2007 at 16:58 GMT
This is a good point.
Indeed, the model’s predictions are dependent on the spatial statistics of the RGC mosaic as well as the coverage ratio (the mean distance between their locations expressed in terms of one-standard deviation of the receptive field center width). As you correctly point out, if these statistics change with eccentricity the model predictions will also depend on eccentricity. This means eccentricity-matched parameters must be used to instantiate the model in particular cases.
Implicit in the question, I think, is the assumption that the cortical architecture is completely uniform. However, this is not the case: there are many examples where RF properties (such as the ratio of simple and complex cells) and map statistics vary with eccentricity. A recent discussion of such work can be found in the recent work of Casagrande and colleagues: Xu et al, J. Comp Neurol, 501:741-55, 2007.
I agree with you that it would be important to study if the instantiation of the model with parameters corresponding to different eccentricities predict some of the trends observed. I have not yet studied this problem.