The skin contains multiple types of sensory nerves that inform the brain about events occurring on the surface of the body. One way to study how this process works is to insert a very fine needle through the skin to stimulate a single sensory nerve with a small electrical current. This technique – known as intraneural microstimulation – can activate touch responses in the brain without an object actually contacting the skin.

Another technique called functional magnetic resonance imaging (fMRI) has been used to measure brain activity. These studies have revealed that when objects come into contact with the skin of the fingers, they stimulate several sensory nerves at the same time, which results in brain activity in a region called the somatosensory cortex.

Sanchez Panchuelo, Ackerley et al. combined fMRI and intraneural microstimulation to map brain activity in response to the activation of individual sensory nerves in the fingers of human volunteers. The experiments show that intraneural stimulation activates many areas of the brain that are also activated by mechanical contact. Future work will use this new method to study the brain's response to signals from different types of sensory nerves.

The primary somatosensory cortex (S1) has been extensively explored in animal studies where it has been shown that this area displays multiple, fine-grained representations of the body (Paul et al., 1972; Kaas et al., 1979; Favorov et al., 1987). Penfield and Boldrey (Penfield & Boldrey 1937) derived the first maps of the somatotopic human body representation in S1 using electrical stimulation of the cortical surface. Somatosensory research in humans has involved using psychophysical (Klatzky et al., 1985; Gescheider et al., 2002), microneurographic (Vallbo & Johansson 1984; Johansson & Vallbo 1983), and neuroimaging (McGlone et al., 2002; Martuzzi et al., 2014; Servos et al., 2001) techniques to study different stages and levels of detail in somatosensory function. Functional magnetic resonance imaging (fMRI) has been used extensively for the non-invasive study of the somatosensory cortices in humans (Nelson and Chen, 2008; McGlone et al., 2002; Sanchez-Panchuelo et al., 2010). Most fMRI studies have investigated the spatial pattern of cortical activation in response to vibrotactile (Francis et al., 2000; Sanchez-Panchuelo et al., 2010) or pneumatic (Huang and Sereno, 2007; Overduin and Servos, 2008) mechanical stimulation of the digits, or to electrical stimulation of the skin (Blankenburg et al., 2003) or median nerve (Kampe et al., 2000; Ferretti et al., 2007). These approaches excite large populations of different classes of mechanoreceptive afferents resulting in relatively diffuse activations in contralateral S1 and bilateral secondary somatosensory cortex (S2).

Microneurography provides a method to record the spike discharge activity of a single mechanoreceptive afferent in conscious humans (Vallbo and Hagbarth, 1968) to determine its response to skin contact and the properties of its receptive field, i.e. location, size, and shape. In this manner, mechanoreceptive afferents innervating the glabrous skin of the hand can be categorized into one of four types: fast-adapting type 1 (FA1), fast-adapting type 2 (FA2), slowly-adapting type 1 (SA1), and slowly-adapting type 2 (SA2) (Vallbo and Johansson, 1984). In intraneural microstimulation (INMS), single mechanoreceptive afferents are selectively activated by passing a small (1–7 μA) current through the recording microelectrode, thus evoking a quantal sensation in the projected sensory field, which matches the physiological qualities of the recorded mechanoreceptive afferent (Torebjörk et al., 1987). Microstimulation of an FA1 afferent evokes a well-defined, local sensation of ‘flutter’ or ‘buzzing’, while microstimulation of an SA1 afferent evokes a sensation of continuous pressure or inward pulling (Vallbo et al., 1984; Ochoa and Torebjörk, 1983). Microstimulation of an FA2 afferent evokes a diffuse sensation of vibration over a larger area, whereas microstimulation of an SA2 afferent does not produce a consistent, conscious sensory experience (Vallbo et al., 1984; Ochoa and Torebjörk, 1983).

It has been shown in a small number of previous studies that INMS of single mechanoreceptive afferents can be combined with noninvasive imaging methods to advance our understanding of the effects of mechanoreceptive afferent activity in somatosensory cortices. For example, INMS of FA1 and SA1 afferents in the median nerve produces frequency-following electroencephalography responses within contralateral S1 (Kelly et al., 1997). The single previous study combining INMS with fMRI (Trulsson et al., 2001), using a 3 T scanner and a surface coil positioned over the parietal lobe contralateral to the site of stimulation, showed that INMS of FA1 and SA1 afferents induced activity in S1 and S2, which overlapped with regions activated by applying mechanical vibration to the relevant units’ receptive fields. However, a detailed characterization of the specificity of single unit INMS activations within the representation of the digits in S1 has yet to be performed.

Several studies have previously assessed the cortical response to vibrotactile stimulation of the glabrous skin of the human hand, and shown that this evokes a hemodynamic response in multiple primary and secondary cortical areas, including contralateral S1, bilateral S2, primary motor cortex (M1), supplementary motor area (SMA), cingulate cortex, posterior parietal cortex (PPC), and insula cortex (McGlone et al., 2002; Trulsson et al., 2001; Gelnar et al., 1998). Ultra-high field (7T) fMRI has also recently been used in conjunction with vibrotactile stimulation to map individual digit representations and resolve the fine, within-digit organization (base-to-tip), thus revealing functional subdivisions of areas in S1 (Sanchez-Panchuelo et al., 2010; Sanchez-Panchuelo et al., 2012). Compared to lower field measurements, 7T fMRI provides greatly increased sensitivity and blood-oxygenation level dependent (BOLD) signal contrast, coupled with improved intrinsic spatial specificity (Gati et al., 1997). Here, we used 7T fMRI to resolve whole-brain cortical activation patterns evoked by INMS of single mechanoreceptive afferent units in the glabrous skin of the hand, and to assess the precise spatial localization of INMS-evoked BOLD responses in contralateral S1, in comparison to activation due to mechanical vibrotactile stimulation.

A total of 33 mechanoreceptive afferents were found (17 FA1, 14 SA1, 1 FA2 and 1 SA2) in 4 participants during 10 experimental sessions. We focused our study on the cortical response to stimulation of type 1 afferents (FA1 and SA1), as these units are far more numerous in the volar hand than type 2 units (FA2 and SA2) (Vallbo and Johansson, 1984). Example recordings from FA1 and SA1 units are shown in Figures 1a and b respectively, demonstrating that good quality signals can be recorded from single mechanoreceptive afferents in the environment of a 7T magnetic resonance scanner. INMS of single units produced distinct sensations: FA1 stimulation was typically felt as vibration or buzzing, while SA1 stimulation elicited a sensation of pressure or pulling (see Table 1).

Figure 1

Download asset Open asset

Due to the technically challenging set-up (e.g. 2 units were lost on moving the participant into the scanner bore) and the nature of the method (e.g. the stimulated unit corresponds to the unit from which recordings were previously made only around 50% of the time [Torebjörk et al., 1987]), INMS was carried out during concurrent fMRI in 11 units (U1-U11) that gave single-point sensations, 6 of which were electrophysiologically-characterized (see Table 1). The receptive field locations for these units are shown in Figure 1c.

Cortical responses to single unit INMS and vibrotactile stimulation in S1

Clear and reproducible BOLD responses were found in somatosensory regions, when INMS was perceived. Occasionally, participants reported that the sensation evoked by the INMS stopped, likely due to a minor dislodgement of the microelectrode. This occurred for U7 where a projected sensation was perceived prior to scanning, but no sensation was felt during the fMRI run. For some units, the sensation was weak (U2, U3; possibly due to difficulty in attending to the stimulus sensation when inside the scanner), or lost during the fMRI run (U5, U6, U8). We compared the location of fMRI responses of all perceived INMS units in contralateral S1 with the digit representation obtained from both vibrotactile stimulation of the microstimulated unit’s receptive field and the fMRI somatotopy maps formed from the traveling-wave (phase-encoding) vibrotactile paradigm (Figure 2). We found that fMRI responses to INMS of single units (all except for U1; Figure 3—figure supplement 1) were spatially localized within the relevant S1 digit representation identified from vibrotactile stimulation. Figure 2a shows example maps of digit somatotopy defined from the vibrotactile traveling-wave paradigm for Participant 4 in the right and left hemispheres (left and right of the figure, respectively). Figure 2b shows the BOLD response to INMS of U11 (right) and U9 (left) for Participant 4. These responses are well-localized within regions of the somatotopic map for digit 4 of the left hand and digit 1 of the right hand, respectively. Figure 2c shows the activation generated in S1 by applying vibrotactile stimulation to the receptive field of U11 (right) and U9 (left). Fits to the hemodynamic responses evoked in S1 by INMS and the application of vibrotactile stimulation to the unit’s receptive field can be seen in Figure 2d.

Figure 2

Download asset Open asset

Figure 3 with 1 supplement see all

Download asset Open asset

Figure 3—source data 1

Source files for plots of Z-score and Proportion of active voxels in each Digit ROI.

This matlab file contains variables for each individual unit (U4, U5, U6, U8, U9, U11) with fields ‘micro_stats’ and ‘vibro_stats’ containing a structure with the results for single unit INMS and vibrotactile stimulation of the unit’s receptive field, respectively. Each structure has the following fields: ‘zetaMean’, ‘betaSem’: (5 digits x 1)-vector containing mean Z-score (FDR-corrected) and standard error across voxels for each Digit ROI; ‘PropActVox’: (5 digits x 1)-vector containing proportion of active voxels (Z>3.08, FDR-corrected) in each Digit ROI; and ‘betaMean’, ‘betaSem’: (5 digits x 1)-vector containing mean GLM parameter estimate and standard error across voxels for each Digit ROI. GLM parameter estimates are not plot in Figure 3 but are used for subsequent group analysis.

https://doi.org/10.7554/eLife.12812.007

Download elife-12812-fig3-data1-v2.zip

Figure 3 shows the spatial localization of the activation produced in S1 by the seven perceived INMS units (U4-U6, U8-U11) (Figure 3a) and corresponding vibrotactile stimulation of each units’ receptive field (Figure 3b). In general, the BOLD responses due to INMS and vibrotactile stimulation were well localized within the expected digit ROI, as defined from the traveling-wave somatotopy paradigm. Figure 3c plots the average INMS z-score (FDR corrected) in each digit ROI, and Figure 3d shows the proportion of active voxels to the INMS paradigm that were classified to each digit ROI (z>3.08, FDR corrected). As expected, the average z-score and proportion of active voxels in the digit ROIs corresponding to digits in which the INMS was sensed was higher than in the neighboring digits. Figure 4 plots the group-level response to show the spatial spread of the INMS and vibrotactile response to neighboring digits. Figure 4a shows the mean z-score, Figure 4b the proportion of active voxels and Figure 4c the GLM parameter estimate to INMS (top) and vibrotactile stimulation of the unit’s receptive field (bottom). ANOVA results showed a significant difference in mean Z-score (F_4,30=14.08, p<10^-5; F_4,30=12.97, p<10^-5), proportion of active voxels (F_4,30=16.12, p<10^-6; F_4,30=17.64, p<10^-6) and GLM parameter estimates (F_4,30=13.52, p<10^-5; F_4,30=14.1, p<10^-5) across the stimulated and neighboring digit classification (INMS; vibrotactile). A multiple pairwise comparison, adjusted for multiple comparisons, showed that measures for the stimulated digit were significantly higher than those of the neighboring digits for mean Z-score (p<0.0001 INMS; p<0.005 vibrotactile stimulation), proportion of active voxels (p<0.00005 for INMS and vibrotactile stimulation) and GLM parameter estimates (p0.01 for INMS and vibrotactile stimulation).

Figure 4

Download asset Open asset

Figure 4—source data 1

Source files for ANOVA tests.

This matlab file contains the 2D-matrices (11 x 5), related to each panel in Figure 4, that were used for the 1-way analysis of variance (performed using the ‘anova1’ matlab command). Each matrix row contains data for each of the 7 units (there are up to eleven 1^st degree neighboring digit ROIs) and each matrix columns represents the ‘proximity’ to the stimulated digit ROI (stimulated digit ROI, 1^st degree, 2^nd degree, 3^rd degree and 4^th degree neighboring digit ROIs). ‘Zeta_micro’ and ‘Zeta_vibro’ are the matrices containing the Z-score (FDR-corrected) values, ‘PerVox_micro’ and ‘PerVox_vibro’ contain the proportion of active voxels (Z>3.08, FDR-corrected) and’ Beta_micro’ and ‘Beta_vibro’ contain the GLM parameter estimates for INMS and vibrotactile stimulation respectively. ANOVA results show a significant difference in mean Z-score (F_4,30=14.08, p<10^-5; F_4,30=12.97, p<10^-5), proportion of active voxels (F_4,30=16.12, p<10^-6; F_4,30=17.64, p<10^-6) and GLM parameter estimates (F_4,30=13.52, p<10^-5; F_4,30=14.1, p<10^-5) across the stimulated and neighboring digit classification (INMS; vibrotactile).

https://doi.org/10.7554/eLife.12812.010

Download elife-12812-fig4-data1-v2.zip

For those units lost during the fMRI run (U5, U6, U8), no areas were found to show a significant correlation with an additional (parametric) regressor when modelling linear reductions in induced response over time (to model gradual losses of unit responses), likely due to the sudden rather than gradual loss of the unit. Thus parameter estimates to INMS stimulation were not significantly different between the GLM including a parametric regressor and modelling the INMS stimulation alone.

Comparison of cortical activity patterns between single unit INMS and vibrotactile stimulation

Participants freely described the mechanical, point-vibrotactile stimulus applied to each unit’s receptive field as feeling very similar in extent and quality to the INMS, especially for the sensations generated from FA1 units. Figure 5a compares the mapping of INMS-induced fMRI responses (yellow) for all FA1 single units to maps of the responses produced by applying vibrotactile stimulation to the units’ receptive fields (blue). Overlapping cortical responses are shown in green. Activation maps show the conjunction of the individual FA1 unit responses, using the same statistical threshold (Z > 3.08, false discovery rate (FDR) correction) for both INMS and vibrotactile stimulation. BOLD responses to single unit INMS were detected in a number of sensory-related brain areas, including S1, S2 (Brodmann areas (BA) 40 and 43), premotor cortex (PMC; SMA and dorsal PMC), M1, insula (anterior insula cortex (AIC) and posterior insula cortex (PIC)), prefrontal cortex (PFC) and PPC. Table 2 details the location and statistical significance (mean and standard error across units) of the BOLD responses produced in these areas by INMS of the five FA1 single units in the left hand. Common areas of activation for INMS and vibrotactile stimulation included S1, S2, PMC, M1, and contralateral PIC; however, INMS gave rise to significant activity in additional brain regions, including the AIC, PPC and contralateral PFC (Table 2). Figure 5b shows that the HRFs generated in these regions by INMS were similar in both onset and duration to the INMS-elicited responses in S1 and S2.

Figure 5

Download asset Open asset

Table 2

			Single unit INMS			Vibrotactile stimulation
ROI	No. Units	x, y, z MNI co-ordinates	Beta	Z	Voxels	Beta	Z	Voxels
SI R SI L	4 3	54, -12, 46 -52, -12, 44	1.4 ± 0.2 1.2 ± 0.2	5.9 ± 0.5 5.6 ± 0.8	38 ± 7 20 ± 9	1.3 ± 0.3 1.6 ± 0.3	5.4 ± 0.3 5.2 ± 0.2	41 ± 12 19 ± 1
BA 40 R BA 40 L	5 4	60, -22, 16 -60, -22, 16	1.4 ± 0.2 1.5 ± 0.4	4.9 ± 0.2 5.3 ± 0.2	56 ± 5 73 ± 5	1.4 ± 0.1 1.4 ± 0.2	4.8 ± 0.2 5.0 ± 0.1	54 ± 7 72 ± 12
BA 43 R BA 43 L	2 3	60, -4, 10 -58, -12, 14	1.1 ± 0.4 1.0 ± 0.4	5.4 ± 0.1 4.8 ± 0.3	45 ± 6 33 ± 8	1.2 ± 0.4 1.7 ± 0.3	4.4 ± 0.2 4.2 ± 0.2	30 ± 20 26 ± 11
SMA R SMA L	5 5	4, 0, 60 -2, 0, 60	1.2 ± 0.2 1.2 ± 0.2	4.8 ± 0.3 4.5 ± 0.3	93 ± 27 66 ± 19	1.3 ± 0.2 1.2 ± 0.1	4.8 ± 0.2 4.5 ± 0.3	43 ± 21 29 ± 6
PMC R PMC L	4 5	54, 0, 50 -52, -2, 50	0.8 ± 0.2 1.1 ± 0.1	4.7 ± 0.2 5.5 ± 0.3	36 ± 11 37 ± 7	1.1 ± 0.2 1.2 ± 0.1	5.0 ± 0.2 4.3 ± 0.1	46 ± 9 20 ± 8
M1 R M1 L	3 2	54, -6, 48 -52, -6, 48	0.9 ± 0.2 1.5 ± 0.2	5.2 ± 0.5 6.3 ± 0.1	51 ± 20 66 ± 36	0.8 ± 0.2 1.3 ± 0.1	5.0 ± 0.7 5.3 ± 0.5	31 ± 10 21 ± 3
PIC R PIC L	5 5	46, -2, 10 -42, -2, 10	0.8 ± 0.2 0.8 ± 0.1	4.2 ± 0.2 4.4 ± 0.2	45 ± 12 38 ± 14	0.8 ± 0.2 -	4.7 ± 0.2 -	27 ± 3 -
AIC R AIC L	4 4	34, 26, 4 -32, 26, 4	1.2 ± 0.1 1.1 ± 0.1	4.7 ± 0.2 4.4 ± 0.2	146 ± 20 106 ± 21	- -	- -	- -
PPC R PPC L	4 5	38, -48, 50 -38, -48, 56	1.2 ± 0.1 1.0 ± 0.1	4.4 ± 0.3 4.4 ± 0.3	168 ± 44 172 ± 43	- -	- -	- -
PFC R	4	42, 34,18	1.2 ± 0.2	4.5 ± 0.3	78 ± 22	-	-	-

Table 2—source data 1

Source files for single unit INMS.

This matlab file contains 2D-matrices (19x5) with the results for single unit INMS for each of the 5 individual units (U1, U4, U6, U8, U11) in each of the 19 ROIs. 'BetaValues’ contains mean across voxels of the beta values, ‘Z-score’ contains the mean Z_score (FDR- corrected) across voxels and ‘NumberVoxels’ contains the number of significant active voxels (Z > 3.08, FDR-corrected) in the ROI. Table 2 summarizes the results by showing the mean and standard error across the 5 units.

https://doi.org/10.7554/eLife.12812.013

Download elife-12812-table2-data1-v2.zip

Table 2—source data 2

Source files for vibrotactile stimulation.

This matlab file contains 2D-matrices (19 ROIs x 5 units) with the results for vibrotactile stimulation applied to the receptive field for each of the 5 individual units (U1, U4, U6, U8, U11) in each ROI. ‘BetaValues ‘contains mean across voxels of the beta values, ‘Z_score’ contains the mean Z-score (FDR- corrected) across voxels and ‘NumberVoxels’ contains the number of significant active voxels (Z > 3.08, FDR-corrected) in the ROI. Table 2 summarizes the results by showing the mean and standard error across the 5 units.

https://doi.org/10.7554/eLife.12812.014

Download elife-12812-table2-data2-v2.zip

The principal finding of our present work is the detailed localization in contralateral S1 of cortical responses to the electrical microstimulation of single, first-order mechanoreceptive afferents, and the demonstration of spatial alignment of these responses with somatotopic maps derived from mechanical skin stimulation. This was achieved through the combined usage of two techniques: intra-neural microstimulation (INMS), to stimulate single mechanoreceptive afferents, and 7T fMRI, to map the cortex with superior spatial resolution. This work also shows that activity generated by stimulation of a single mechanoreceptive afferent can be perceptually characterized and produces a network of cortical responses.

Only one previous study has combined single unit INMS with fMRI, at 3T (Trulsson et al., 2001), but this was only able to resolve activation in contralateral S1 and S2 as the use of a surface coil limited the spatial extent of activation maps. The greater signal-to-noise ratio and improved BOLD contrast afforded by 7T fMRI allowed us to improve the spatial resolution, with a reduction in the voxel volume by a factor of 6 compared to previous work at 3T (Trulsson et al., 2001). We have exploited the improved spatial resolution to provide a detailed characterization of the location and extent of the cortical network involved in encoding inputs from single mechanoreceptive afferents, as well as in comparing these responses to somatotopical maps created from vibrotactile skin stimulation.

Measurements of cortical activity elicited by INMS demonstrated that when a singular, quantal touch from the stimulation of a single mechanoreceptive afferent is consciously felt, a precise area in contralateral S1 is active. The response in S1 was well-localized within the expected region, identified from maps of digit somatotopy obtained from vibrotactile stimulation of the fingertips. The extent of the S1 responses to INMS was less than that elicited by vibrotactile stimulation to the unit’s receptive field, although the response produced by single unit INMS was relatively extensive, considering that vibrotactile stimulation simultaneously engages a large number of afferents (Johansson and Vallbo, 1979; Vallbo and Johansson, 1984).

Robust responses were found within the expected digital cortical area for all perceived microstimulated afferents (Figures 2 and 3), except for U1, for which no significant responses were found, in either contralateral or ipsilateral S1, despite the fact that the participant exhibited a complete somatotopic map of the digits in both hemispheres and reported feeling the sensation throughout INMS. To explore this finding further, we used the delineation of digit 2 from the somatotopic map obtained with the vibrotactile traveling-wave paradigm to inspect the time series of S1 responses evoked by INMS for U1 (located on the palm below digit 2). We also interrogated the BOLD response produced in contralateral S1 when vibrotactile stimulation was applied to the receptive fields of U1. In S1, we found negative BOLD responses (Figure 3—figure supplement 1) for both INMS and vibrotactile stimulation applied to the receptive field of the INMS. The negative BOLD response in this subject is possibly due to a steal effect from the nearby vasculature draining from the active cortex (Bianciardi et al., 2011) since draining venous regions are highly modulated by block paradigms with periods of 'on’ and ‘off’ stimulation, as used to study the response to INMS and vibrotactile stimulation of the receptive field. In contrast, using the traveling-wave paradigm a complete map of the digits in S1 is seen. This is expected, as we have previously shown that a traveling-wave design is insensitive to the non-specific BOLD contributions from large veins that drain blood from across the whole hand representation in S1 (Uğurbil et al., 2003; Besle et al., 2013), thus suppressing the venous signal modulations found in the block INMS/vibrotactile stimulation data. In order to estimate the spatial spread of INMS BOLD responses to neighboring digits, we show that, at the group level, the z-score, proportion of active voxels and GLM parameter estimates are significantly higher (p<0.01) in the stimulated ROI than in the neighboring digits (Figure 4). These results are in-line with our previous findings reported for vibrotactile stimulation (Besle et al.,2013).

The network of cortical areas activated by both INMS of single mechanoreceptive afferents and mechanical vibrotactile stimulation of the units’ receptive field, included somatosensory areas such as S1, S2, and PIC, as well as areas involved in motor control, including M1, SMA and PMC. Although M1 has previously been shown to be activated by tactile input (e.g. Francis et al., 2000; Ackerley et al., 2012), we cannot exclude the possibility that the M1 activation observed in this study may originate from spatial blurring of somatosensory activation (given that M1 and S1 are located on opposite banks of the central sulcus). When comparing responses to INMS and vibrotactile stimulation applied to the afferents’ receptive fields, INMS activated a number of additional areas, specifically the AIC, PPC and PFC. Exploration of the INMS BOLD time series for these areas (Figure 5b) suggests that the activity in these areas is locked to the S1/S2 activity and is not due to anticipation. Both insula and parietal cortices have been shown to contribute to the perception of touch (Preusser et al., 2014), and a previous study of tactile attention (Burton et al., 2008) has shown that a fronto-parietal network, which includes PFC and PPC, is involved in attention. Although identical paradigm timings were used for INMS and vibrotactile stimulation in order to compare the spatial localization of the BOLD response, there were differences in the attentional focus between the INMS and vibrotactile tasks. During the INMS fMRI runs, participants were aware that perception might be lost and hence had to concentrate on the stimulus and report any lack of sensation at the end of the run. In contrast, the vibrotactile stimulus was delivered at a suprathreshold level and participants did not have to monitor that the sensation was still present during the vibrotactile fMRI run. Hence, the increased activity in AIC, PFC and PPC observed in the present study may reflect the increased attentional effects (i.e., baseline or gain effects on evoked responses) during the INMS protocol compared to vibrotactile stimulation. However, this is a preliminary finding and requires further investigation with larger sample sizes and more quantitative analysis to be corroborated.

The capability of combining INMS with 7T fMRI has the following theoretical implications for human somatosensory research. Although the notion that peripheral input from the skin is represented directly by four cytoarchitectonic areas (BA 3a, 3b, 1 and 2) in S1, each containing an orderly somatotopic map of the body surface has been supported by findings from animal studies (Kaas et al., 1979; Paul et al., 1972; Favorov et al., 1987; Tommerdahl et al., 2010) and 7T fMRI in humans (Sanchez-Panchuelo et al., 2010; Sanchez-Panchuelo et al., 2012; Martuzzi et al., 2014), a simple point-to-point topographical correspondence between skin surface and cortical representation does not hold. In reality, there is integration and processing through axonal synapsing in the dorsal column nuclei and thalamus prior to mechanoreceptive information entering the cerebral cortex. There appears to be a preserved transmission from single, mechanoreceptive second-order neurons in the dorsal column (Vickery et al., 1994). At the level of the thalamus, an axon of a single ventral posterolateral nucleus terminates over a fairly wide, roughly 0.5 mm, cortical territory (Rausell and Jones, 1995), where many individual thalamocortical axons spread out in discrete patches over several millimeters of S1 (Landry et al., 1987). This spread corresponds well with our finding that the cortical activation from a single mechanoreceptive afferent extends over an area that is not dissimilar to the area activated by input from many afferents through point-vibrotactile stimulation. Also, neurons in S1 cortical columns have extensive lateral excitatory connections, not only with neighboring neurons, but also with neurons several millimeters away in the same cortical area (Burton and Fabri, 1995). We have shown that single unit INMS produces bilateral somatosensory activation, as well as influencing motor areas and cognitive networks (e.g. PPC, PFC). Such a wide spreading of stimulus-evoked activity has been clearly documented in microelectrode recording studies (Reed et al., 2010). Overall, the spatiotemporal pattern of S1 response to vibrotactile stimulation is far from simple and its functional significance remains to be unraveled.

Translational insights from in vivo neurophysiological studies in non-human primates have driven much of the theoretical understanding of cortical mechanisms that govern human tactile perception, but operative procedures, especially those which alter the neurochemistry of cortical synaptic transmission (Masamoto et al., 2009), may confound relating such findings to normal functioning of the human brain. This demonstration of the feasibility of combining INMS with 7T fMRI opens up the possibility of a range of further neuroimaging studies that will allow interrogation of the precise anatomical and physiological properties of the fundamental encoding of touch. These include systematic investigation of the sub-cortical (e.g. thalamic) responses and laminar-specific cortical responses to INMS of different mechanoreceptive afferent classes using a variety of electrical stimulation patterns.

Ten experimental sessions were conducted on four right-handed participants (30–64 years, 2 male). Procedures were approved by the University of Nottingham Medical School Ethics Committee and all participants gave full, written, informed consent. Due to the precision needed in performing INMS within the magnetic resonance scanner, participants were required to lie extremely still and feel relaxed; all participants were accustomed to the fMRI environment (two participants had participated in INMS experiments previously). Each experimental session involved three steps: (1) microneurography for the characterization of a single mechanoreceptive afferent (Vallbo and Hagbarth, 1968); (2) assessment of the sensation to INMS; (3) concurrent INMS and fMRI. Participants subsequently took part in a second fMRI session in which vibrotactile stimulation was delivered.

Participants lay on the scanner bed with their arm (the left arm in all cases except one experiment on the right arm) immobilized using cushions. Survey, reference and B₀-map scans were acquired, and an image-based shimming approach (Sanchez-Panchuelo et al., 2010) used to minimize magnetic field inhomogeneity, with the optimized shim currents remaining fixed throughout the subsequent fMRI runs. The participant was moved out of the bore of the magnet to perform Steps (1) and (2).

Microneurography

Request a detailed protocol

In Step 1, the median nerve was accessed at the wrist in order to isolate single axonal responses from mechanoreceptive afferents in the volar hand, on which to perform INMS (Trulsson et al., 2001). A high-impedance (~300–500 kΩ), insulated, tungsten recording/stimulating electrode (15 mm length, shaft diameter 0.2 mm, tip diameter ~5 µm; FHC, Bowdoin, ME) was inserted percutaneously into the skin, ~3 cm from the wrist fold between the flexor carpi radialis and the flexor palmaris longus tendons. An uninsulated reference electrode was inserted subcutaneously 3–5 cm away, on the ulnar side of the recording/stimulating electrode, and a ground electrode was attached further up the participant’s arm (Figure 6). The recording/stimulating electrode was advanced into the median nerve, which was located 0.3–1 cm below the skin surface. The preamplifier was taped to the participant’s arm, and the acquisition hardware and stimulator were located at the outer edge of the scanner room (Figure 6). Differential responses were amplified (x10,000) using a preamplifier (NeuroAmpEX; ADInstruments, Castle Hill, Australia), band-pass filtered (0.3–5 kHz) and sampled at 10 kHz using PowerLab hardware and LabChart 7 software (ADInstruments, Castle Hill, Australia).

Figure 6

Download asset Open asset

The microneurographer delivered light, stroking touch to the palm to evoke activity in low-threshold mechanoreceptive afferents. A loudspeaker in the scanner room allowed the microneurographer to hear the nerve activity and a projector displayed the recording onto the scanner exterior for visual inspection. The microneurographer systematically searched for the nerve until modulations of the signal from the electrode corresponded to mass activity from mechanoreceptive afferents as a result of touch were heard. Using fine adjustments, the electrode was manipulated within the nerve to an intra-fascicular location and single units were searched for by stroking the participant’s hand.

Single mechanoreceptive afferents were characterized by their audio and visual signals, and the extent of the receptive field of each afferent was explored using a wooden stick. The location of the receptive field was mapped using von Frey monofilaments and the minimal force required for mechanoreceptor activation noted. Afferents were identified as being myelinated Aβ mechanoreceptors, namely FA1, SA1, FA2 or SA2 afferents (Vallbo and Johansson, 1984). The middle of the receptive field was marked on the skin. Recordings of individual mechanoreceptive afferents in response to mechanical stimulation were made (e.g. Figure 1a,b) and analyzed in MATLAB (The Mathworks; Natick, MA). Data were preprocessed to verify the single-unit nature of all recorded mechanoreceptive afferents with an offline pattern-matching algorithm.

1. 1. Sanchez Panchuelo RM
   2. Ackerley R
   3. Glover PM
   4. Bowtell RW
   5. Wessberg J
   6. Francis ST
   7. McGlone F

   (2016) fMRI to intraneural microstimulation of single mechanoreceptive afferents

   Available at figshare under a CC0 Public Domain licence.

   https://dx.doi.org/10.6084/m9.figshare.3205552

1. 1. Ackerley R
   2. Hassan E
   3. Curran A
   4. Wessberg J
   5. Olausson H
   6. McGlone F

   (2012) An fMRI study on cortical responses during active self-touch and passive touch from others

   Frontiers in Behavioral Neuroscience 6:.

   https://doi.org/10.3389/fnbeh.2012.00051

   • Google Scholar
2. 1. Benjamini Y
   2. Hochberg Y

   (1995) Controlling the false discovery rate: A practical and powerful approach to multiple testing

   Journal of the Royal Statistical Society. Series B 57:289–300.

   https://doi.org/10.2307/2346101

   • Google Scholar
3. 1. Benjamini Y
   2. Krieger AM
   3. Yekutieli D

   (2006) Adaptive linear step-up procedures that control the false discovery rate

   Biometrika 93:491–507.

   https://doi.org/10.1093/biomet/93.3.491

   • Google Scholar
4. 1. Besle J
   2. Sánchez-Panchuelo RM
   3. Bowtell R
   4. Francis S
   5. Schluppeck D

   (2013) Single-subject fmri mapping at 7 T of the representation of fingertips in S1: A comparison of event-related and phase-encoding designs

   Journal of Neurophysiology 109:2293–2305.

   https://doi.org/10.1152/jn.00499.2012

   • Google Scholar
5. 1. Bianciardi M
   2. Fukunaga M
   3. van Gelderen P
   4. de Zwart JA
   5. Duyn JH

   (2011) Negative bold-fmri signals in large cerebral veins

   Journal of Cerebral Blood Flow and Metabolism 31:401–412.

   https://doi.org/10.1038/jcbfm.2010.164

   • Google Scholar
6. 1. Blankenburg F
   2. Ruben J
   3. Meyer R
   4. Schwiemann J
   5. Villringer A

   (2003) Evidence for a rostral-to-caudal somatotopic organization in human primary somatosensory cortex with mirror-reversal in areas 3b and 1

   Cerebral Cortex 13:987–993.

   https://doi.org/10.1093/cercor/13.9.987

   • Google Scholar
7. 1. Burton H
   2. Fabri M

   (1995) Ipsilateral intracortical connections of physiologically defined cutaneous representations in areas 3b and 1 of macaque monkeys: Projections in the vicinity of the central sulcus

   The Journal of Comparative Neurology 355:508–538.

   https://doi.org/10.1002/cne.903550404

   • Google Scholar
8. 1. Burton H
   2. Sinclair RJ
   3. McLaren DG

   (2008) Cortical network for vibrotactile attention: A fmri study

   Human Brain Mapping 29:207–221.

   https://doi.org/10.1002/hbm.20384

   • Google Scholar
9. 1. Favorov OV
   2. Diamond ME
   3. Whitsel BL

   (1987) Evidence for a mosaic representation of the body surface in area 3b of the somatic cortex of cat

   Proceedings of the National Academy of Sciences of the United States of America 84:6606–6610.

   https://doi.org/10.1073/pnas.84.18.6606

   • Google Scholar
10. 1. Ferretti A
    2. Babiloni C
    3. Arienzo D
    4. Del Gratta C
    5. Rossini PM
    6. Tartaro A
    7. Romani GL

    (2007) Cortical brain responses during passive nonpainful median nerve stimulation at low frequencies (0.5-4 hz): An fmri study

    Human Brain Mapping 28:645–653.

    https://doi.org/10.1002/hbm.20292

    • Google Scholar
11. 1. Francis ST
    2. Kelly EF
    3. Bowtell R
    4. Dunseath WJR
    5. Folger SE
    6. McGlone F

    (2000) fMRI of the responses to vibratory stimulation of digit tips

    NeuroImage 11:188–202.

    https://doi.org/10.1006/nimg.2000.0541

    • Google Scholar
12. 1. Gati JS
    2. Menon RS
    3. Ugurbil K
    4. Rutt BK

    (1997) Experimental determination of the BOLD field strength dependence in vessels and tissue

    Magnetic Resonance in Medicine 38:296–302.

    https://doi.org/10.1002/mrm.1910380220

    • Google Scholar
13. 1. Gelnar PA
    2. Krauss BR
    3. Szeverenyi NM
    4. Apkarian AV

    (1998) Fingertip representation in the human somatosensory cortex: An fmri study

    NeuroImage 7:261–283.

    https://doi.org/10.1006/nimg.1998.0341

    • Google Scholar
14. 1. Gescheider GA
    2. Bolanowski SJ
    3. Pope JV
    4. Verrillo RT

    (2002) A four-channel analysis of the tactile sensitivity of the fingertip: Frequency selectivity, spatial summation, and temporal summation

    Somatosensory & Motor Research 19:114–124.

    https://doi.org/10.1080/08990220220131505

    • Google Scholar
15. 1. Huang R-S
    2. Sereno MI

    (2007) Dodecapus: An mr-compatible system for somatosensory stimulation

    NeuroImage 34:1060–1073.

    https://doi.org/10.1016/j.neuroimage.2006.10.024

    • Google Scholar
16. 1. Hurley AC
    2. Al-Radaideh A
    3. Bai L
    4. Aickelin U
    5. Coxon R
    6. Glover P
    7. Gowland PA

    (2010) Tailored RF pulse for magnetization inversion at ultrahigh field

    Magnetic Resonance in Medicine 63:51–58.

    https://doi.org/10.1002/mrm.22167

    • Google Scholar
17. 1. Johansson RS
    2. Vallbo AB

    (1979) Tactile sensibility in the human hand: Relative and absolute densities of four types of mechanoreceptive units in glabrous skin

    The Journal of Physiology 286:283–300.

    https://doi.org/10.1113/jphysiol.1979.sp012619

    • Google Scholar
18. 1. Johansson RS
    2. Vallbo Åke B

    (1983) Tactile sensory coding in the glabrous skin of the human hand

    Trends in Neurosciences 6:27–32.

    https://doi.org/10.1016/0166-2236(83)90011-5

    • Google Scholar
19. 1. Kaas J
    2. Nelson R
    3. Sur M
    4. Lin C
    5. Merzenich M

    (1979) Multiple representations of the body within the primary somatosensory cortex of primates

    Science 204:521–523.

    https://doi.org/10.1126/science.107591

    • Google Scholar
20. 1. Kampe KK
    2. Jones RA
    3. Auer DP

    (2000) Frequency dependence of the functional MRI response after electrical median nerve stimulation

    Human Brain Mapping 9:106–114.

    https://doi.org/10.1002/(SICI)1097-0193(200002)9:2<106::AID-HBM5>3.0.CO;2-Y

    • Google Scholar
21. 1. Kelly EF
    2. Trulsson M
    3. Folger SE

    (1997)

    Periodic microstimulation of single mechanoreceptive afferents produces frequency-following responses in human EEG

    Journal of Neurophysiology 77:137–144.

    • Google Scholar
22. 1. Klatzky RL
    2. Lederman SJ
    3. Metzger VA

    (1985) Identifying objects by touch: An "expert system"

    Perception & Psychophysics 37:299–302.

    https://doi.org/10.3758/BF03211351

    • Google Scholar
23. 1. Landry P
    2. Diadori P
    3. Leclerc S
    4. Dykes RW

    (1987) Morphological and electrophysiological characteristics of somatosensory thalamocortical axons studied with intra-axonal staining and recording in the cat

    Experimental Brain Research 65:317–330.

    https://doi.org/10.1007/BF00236304

    • Google Scholar
24. 1. Martuzzi R
    2. van der Zwaag W
    3. Farthouat J
    4. Gruetter R
    5. Blanke O

    (2014) Human finger somatotopy in areas 3b, 1, and 2: A 7T fmri study using a natural stimulus

    Human Brain Mapping 35:213–226.

    https://doi.org/10.1002/hbm.22172

    • Google Scholar
25. 1. Masamoto K
    2. Fukuda M
    3. Vazquez A
    4. Kim SG

    (2009) Dose-dependent effect of isoflurane on neurovascular coupling in rat cerebral cortex

    The European Journal of Neuroscience 30:242–250.

    https://doi.org/10.1111/j.1460-9568.2009.06812.x

    • Google Scholar
26. 1. McGlone F
    2. Kelly EF
    3. Trulsson M
    4. Francis ST
    5. Westling G
    6. Bowtell R

    (2002) Functional neuroimaging studies of human somatosensory cortex

    Behavioural Brain Research 135:147–158.

    https://doi.org/10.1016/S0166-4328(02)00144-4

    • Google Scholar
27. 1. Nelson AJ
    2. Chen R

    (2008) Digit somatotopy within cortical areas of the postcentral gyrus in humans

    Cerebral Cortex (New York, N.Y. : 1991) 18:2341–2351.

    https://doi.org/10.1093/cercor/bhm257

    • Google Scholar
28. 1. Ochoa J
    2. Torebjörk E

    (1983) Sensations evoked by intraneural microstimulation of single mechanoreceptor units innervating the human hand

    The Journal of Physiology 342:633–654.

    https://doi.org/10.1113/jphysiol.1983.sp014873

    • Google Scholar
29. 1. Overduin SA
    2. Servos P
    3. &
    4. Servos P

    (2008) Symmetric sensorimotor somatotopy

    PloS One 3:e1505.

    https://doi.org/10.1371/journal.pone.0001505

    • Google Scholar
30. 1. Paul RL
    2. Merzenich M
    3. Goodman H

    (1972) Representation of slowly and rapidly adapting cutaneous mechanoreceptors of the hand in Brodmann’s areas 3 and 1 of Macaca mulatta

    Brain Research 36:229–249.

    https://doi.org/10.1016/0006-8993(72)90732-9

    • Google Scholar
31. 1. Penfield W
    2. Boldrey E

    (1937) SOMATIC motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation

    Brain 60:389–443.

    https://doi.org/10.1093/brain/60.4.389

    • Google Scholar
32. 1. Preusser S
    2. Thiel SD
    3. Rook C
    4. Roggenhofer E
    5. Kosatschek A
    6. Draganski B
    7. Blankenburg F
    8. Driver J
    9. Villringer A
    10. Pleger B

    (2015) The perception of touch and the ventral somatosensory pathway

    Brain 138:.

    https://doi.org/10.1093/brain/awu370

    • Google Scholar
33. 1. Rausell E
    2. Jones EG

    (1995)

    Extent of intracortical arborization of thalamocortical axons as a determinant of representational plasticity in monkey somatic sensory cortex

    Journal of Neuroscience  15:4270–4288.

    • Google Scholar
34. 1. Reed JL
    2. Qi HX
    3. Pouget P
    4. Burish MJ
    5. Bonds AB
    6. Kaas JH

    (2010) Modular processing in the hand representation of primate primary somatosensory cortex coexists with widespread activation

    Journal of Neurophysiology 104:3136–3145.

    https://doi.org/10.1152/jn.00566.2010

    • Google Scholar
35. 1. Sanchez-Panchuelo RM
    2. Francis S
    3. Bowtell R
    4. Schluppeck D

    (2010) Mapping human somatosensory cortex in individual subjects with 7T functional MRI

    Journal of Neurophysiology 103:2544–2556.

    https://doi.org/10.1152/jn.01017.2009

    • Google Scholar
36. 1. Sanchez-Panchuelo RM
    2. Besle J
    3. Beckett A
    4. Bowtell R
    5. Schluppeck D
    6. Francis S

    (2012) Within-digit functional parcellation of brodmann areas of the human primary somatosensory cortex using functional magnetic resonance imaging at 7 tesla

    Journal of Neuroscience 32:15815–15822.

    https://doi.org/10.1523/JNEUROSCI.2501-12.2012

    • Google Scholar
37. 1. Servos P
    2. Lederman S
    3. Wilson D
    4. Gati J

    (2001) fMRI-derived cortical maps for haptic shape, texture, and hardness

    Brain Research. Cognitive Brain Research 12:307–313.

    https://doi.org/10.1016/s0926-6410(01)00041-6.

    • Google Scholar
38. 1. Tommerdahl M
    2. Favorov OV
    3. Whitsel BL

    (2010) Dynamic representations of the somatosensory cortex

    Neuroscience and Biobehavioral Reviews 34:160–170.

    https://doi.org/10.1016/j.neubiorev.2009.08.009

    • Google Scholar
39. 1. Torebjörk HE
    2. Vallbo AB
    3. Ochoa JL

    (1987) Intraneural microstimulation in man. its relation to specificity of tactile sensations

    Brain : A Journal of Neurology 110 ( Pt 6:1509–1529.

    https://doi.org/10.1093/brain/110.6.1509

    • Google Scholar
40. 1. Trulsson M
    2. Francis ST
    3. Kelly EF
    4. Westling G
    5. Bowtell R
    6. McGlone F

    (2001) Cortical responses to single mechanoreceptive afferent microstimulation revealed with fmri

    NeuroImage 13:613–622.

    https://doi.org/10.1006/nimg.2000.0723

    • Google Scholar
41. 1. Uğurbil K
    2. Toth L
    3. Kim DS

    (2003) How accurate is magnetic resonance imaging of brain function?

    Trends in Neurosciences 26:108–114.

    https://doi.org/10.1016/S0166-2236(02)00039-5

    • Google Scholar
42. 1. Vallbo AB
    2. Hagbarth KE

    (1968) Activity from skin mechanoreceptors recorded percutaneously in awake human subjects

    Experimental Neurology 21:270–289.

    https://doi.org/10.1016/0014-4886(68)90041-1

    • Google Scholar
43. 1. Vallbo AB
    2. Johansson RS

    (1984)

    Properties of cutaneous mechanoreceptors in the human hand related to touch sensation

    Human Neurobiology 3:3–14.

    • Google Scholar
44. 1. Vickery RM
    2. Gynther BD
    3. Rowe MJ

    (1994) Synaptic transmission between single slowly adapting type I fibres and their cuneate target neurones in cat

    The Journal of Physiology 474:379–392.

    https://doi.org/10.1113/jphysiol.1994.sp020030

    • Google Scholar
45. 1. Vallbo AB
    2. Olsson KA
    3. Westberg KG
    4. Clark FJ

    (1984) Microstimulation of single tactile afferents from the human hand. Sensory attributes related to unit type and properties of receptive fields

    Brain : A Journal of Neurology 107:727–749.

    https://doi.org/10.1093/brain/107.3.727

    • Google Scholar

Author details

1. Rosa Maria Sanchez Panchuelo

   Sir Peter Mansfield Imaging Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, United Kingdom

   Contribution

   RMSP, Contributed to the design of the study, Developed the magnetic resonance sequences and acquired the MRI data, Performed all the fMRI/MRI data analysis, Contributed to the interpretation of the results and drafted the paper

   Contributed equally with

   Rochelle Ackerley

   For correspondence

   rosa.panchuelo@nottingham.ac.uk

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-9917-0020

2. Rochelle Ackerley

   1. Department of Physiology, University of Gothenburg, Göteborg, Sweden
   2. Laboratoire de Neurosciences Intégratives et Adaptatives, Aix-Marseille University, Marseille, France

   Contribution

   RA, Contributed to the design of the study, Conducted the microneurography and analyzed the microneurography data, Contributed to the interpretation of the results and drafting of the paper

   Contributed equally with

   Rosa Maria Sanchez Panchuelo

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0003-4621-7929

3. Paul M Glover

   Sir Peter Mansfield Imaging Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, United Kingdom

   Contribution

   PMG, Contributed to the design of the study and collection of the data, Adapted the standard microneurography equipment for compatibility with an ultra-high field MRI scanner, Contributed to the interpretation of the results and revised the paper

   Competing interests

   The authors declare that no competing interests exist.

4. Richard W Bowtell

   Sir Peter Mansfield Imaging Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, United Kingdom

   Contribution

   RWB, Contributed to the design of the study and collection of the data, Contributed to the interpretation of the results and revised the paper

   Competing interests

   The authors declare that no competing interests exist.

5. Johan Wessberg

   Department of Physiology, University of Gothenburg, Göteborg, Sweden

   Contribution

   JW, Contributed to the design of the study and collection of the data, Conducted the microneurography, Contributed to the interpretation of the results and revised the paper

   Competing interests

   The authors declare that no competing interests exist.

6. Susan T Francis

   Sir Peter Mansfield Imaging Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, United Kingdom

   Contribution

   STF, Contributed to the design of the study, Developed the magnetic resonance sequences and acquired the MRI data, Guided the fMRI analysis and contributed to the interpretation of the results and writing of the paper

   Contributed equally with

   Francis McGlone

   Competing interests

   The authors declare that no competing interests exist.

7. Francis McGlone

   School of Natural Sciences and Psychology, Liverpool John Moores University, Liverpool, United Kingdom

   Present address

   Institute of Psychology, Health and Society, University of Liverpool, Liverpool, United Kingdom

   Contribution

   FM, Contributed to the design of the study and collection of the data, Contributed to the interpretation of the results and writing of the paper

   Contributed equally with

   Susan T Francis

   Competing interests

   The authors declare that no competing interests exist.

• 1,332

  views

• 351

  downloads

• 33

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.