The organizational principles of de-differentiated topographic maps in somatosensory cortex

Abstract

<jats:title>Abstract</jats:title><jats:p>Topographic maps are a fundamental feature of cortex architecture in the mammalian brain. One common theory is that the de-differentiation of topographic maps links to impairments in everyday behavior due to less precise functional map readouts. Here, we tested this theory by characterizing de-differentiated topographic maps in primary somatosensory cortex (SI) of younger and older adults by means of ultra-high resolution functional magnetic resonance imaging together with perceptual finger individuation and hand dexterity. Older adults’ SI maps showed similar amplitude, size, and levels of stimulus-related noise than younger adults’ SI maps, but presented with less representational similarity between distant fingers. Larger population receptive field sizes in older adults’ maps did not correlate with behavior, whereas reduced cortical distances related to better hand dexterity. Our data uncover the drawbacks of a simple de-differentiation model of topographic map function, and motivate the introduction of a feature-based model of cortical reorganization.</jats:p>

Fast event-related mapping of population fingertip tuning properties in human sensorimotor cortex at 7T

“…Performing a task might increase integration between information from different fingertips. However, pRF width estimates in these two studies are much lower than those from two previous studies (Liu et al, 2021; Puckett et al, 2020) that reported fingertip pRF widths of ∼9-10 fingertips FWHM on average across S1. Since one study used passive stimulation (Puckett et al, 2020) and the other used an active task (Liu et al, 2021), task differences cannot account for these larger pRF width estimates.…”

“…However, pRF width estimates in these two studies are much lower than those from two previous studies (Liu et al, 2021; Puckett et al, 2020) that reported fingertip pRF widths of ∼9-10 fingertips FWHM on average across S1. Since one study used passive stimulation (Puckett et al, 2020) and the other used an active task (Liu et al, 2021), task differences cannot account for these larger pRF width estimates. Another possible explanation is that the two latter studies derived pRF parameters from phase-encoding stimulation sequences whereas both we and Schellekens et al (2021) used pseudo-random stimulation sequences that included periods without stimulation (better suited to estimate pRF width in regions with large receptive fields because they allow the BOLD response to return to baseline; Dumoulin and Wandell, 2008).…”

“…This is statistically similar to what was reported by Puckett et al (2020), who reported similar pRF widths for the index/middle/ring representations (the thumb was not stimulated). While the increase in pRF width from thumb to little finger was monotonic in BA1, our data suggest a possible local pRF width maximum for the index fingertip representation in BA3b (not significant), similar to that reported in S1 by Liu et al (2021) and found to be significant in their study. Therefore at this point, it is difficult to say whether pRF width increases monotonically from thumb to little finger representation or whether the pattern is more complex, at least in some BAs.…”

“…The smaller pre-central somatotopic representation will be discussed in more detail in section 4.2 of the discussion. Fingertip tuning curves have previously been measured only in somatotopically-organized regions, either by averaging tuning curves across voxels in regions responding preferentially to the same fingertip (Besle et al, 2014, 2013; Martuzzi et al, 2014; Stringer et al, 2014; van der Zwaag et al, 2015), or by fitting Gaussian fingertip pRF models voxelwise (Liu et al, 2021; Puckett et al, 2020; Schellekens et al, 2018, 2021). Fitting pRF models voxelwise also allows the tuning width to be measured in regions that are not somatotopically-organized.…”

“…While phase-encoding designs are very efficient at mapping the dominant body part in a given cortical location, block and ER designs allow for the independent estimation of BOLD responses to the stimulation of different body parts in the same cortical location (Besle et al, 2014). This makes it possible to study how the cortical representations of different body parts overlap (Besle et al, 2014; Kolasinski et al, 2016b; Liu et al, 2021) and to measure local selectivity to different body parts (Besle et al, 2014; Martuzzi et al, 2014; Stringer et al, 2014), including the estimation of Gaussian population receptive fields (pRF) at the voxel level (Liu et al, 2021; Puckett et al, 2020; Saadon-Grosman et al, 2020a; Schellekens et al, 2018, 2021). Compared with ER designs, block designs are statistically more powerful at detecting differences between experimental conditions (Birn et al, 2002; Friston et al, 1999) and are therefore better suited if scanning time is a limiting factor.…”