High-resolution maps of real and illusory tactile activation in primary somatosensory cortex in individual monkeys with functional magnetic resonance imaging and optical imaging - PubMed (original) (raw)

High-resolution maps of real and illusory tactile activation in primary somatosensory cortex in individual monkeys with functional magnetic resonance imaging and optical imaging

Li M Chen et al. J Neurosci. 2007.

Abstract

Although blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) has been widely used to explore human brain function, questions remain regarding the ultimate spatial resolution of positive BOLD fMRI, and indeed the extent to which functional maps revealed by positive BOLD correlate spatially with maps obtained with other high-spatial-resolution mapping techniques commonly used in animals, such as optical imaging of intrinsic signal (OIS) and single-unit electrophysiology. Here, we demonstrate that the positive BOLD signal at 9.4T can reveal the fine topography of individual fingerpads in single-condition activation maps in nonhuman primates. These digit maps are similar to maps obtained from the same animal using intrinsic optical imaging. Furthermore, BOLD fMRI reliably resolved submillimeter spatial shifts in activation in area 3b previously identified with OIS (Chen et al., 2003) as neural correlates of the "funneling illusion." These data demonstrate that at high field, high-spatial-resolution topographic maps can be achieved using the positive BOLD signal, weakening previous notions regarding the spatial specificity of the positive BOLD signal.

PubMed Disclaimer

Figures

Figure 1.

Anatomic images for studying SI. a, A high-resolution coronal image is collected to locate somatosensory cortices and to guide placement of three oblique slices parallel to the SI (locations indicated by rectangles, overlaid on coronal scout image). This oblique orientation was used for both high-resolution anatomical and functional imaging. b, Major landmarks (such as central and lateral sulci) used to identify SI are visible on squirrel monkey brain. c, In three images acquired with T2* weighting, sulci and vascular structures appear dark. Both central and lateral sulci are readily identified in the most superficial slice. Scale bar, 10 mm. a, Anterior; m, medial.

Figure 2.

fMRI of digit topography in two monkeys. a–c, f, g, Single-condition activation maps of vibrotactile stimulation of D4 (a), D3 (b), and D2 (c) in monkey 1, and D2 (f) and D1 (g) in monkey 2. The significance levels of the activation-associated changes in BOLD signal in a–c and f and g are indicated by their respective color bars and corresponding p value ranges. Focal activation is observed in area 3b in all five images, and area 1 activation is also apparent in a and g. d, h, Composite maps of a–c in monkey 1 and f–g in monkey 2, respectively (different colors represent different digit activations: D4, blue; D3, green; D2, red; D1, orange). Two inserts show the enlarged view of the composite fingerpad maps. Scale bars: (in d) a–d, (in h) f–h, 5 mm; insets, 1 mm. e, Fingerpad topography in area 3b closely matches the known somatotopic map of squirrel monkey area 3b and area 1 [from Sur et al., (1982)]. i, The raw signal time course (signal amplitude in scanner arbitrary units) for a single run using block design (five cycles of 20 frames of stimulus/20 frames of baseline, corresponding to five 30 s blocks of stimulus and baseline) from D4 activation in area 3b (a). j, Time course of mean percentage signal for five blocks from the same run. The mean percentage signal changes (±SD; k) were highly reproducible and stable across three different sessions (days) in the same monkey (left), across three runs within the same session (same day) in one monkey (middle), and across seven different monkeys (right). In monkey 1, activations of D4 (a) and D3 (b) are thresholded at p < 10−7 and D2 at p < 10−8. In monkey 2, activations of D1 and D2 are thresholded at p < 10−8. a, Anterior; m, medial.

Figure 3.

Effect of different image thresholds on the location and size of BOLD fMRI activation. a–h, Area of D1 activation in case 1 (a–c) and D3 activation in case 2 (e–g) decrease with increasing stringency of thresholding (p < 10−4; p < 5 × 10−5; and p < 10−5), but composite overlays (d, h) confirm that essential topography and the centroid of activation are unchanged. i, The activation area increases with increasing p value threshold for D1 in case 1 and D3 in case 2 (area relative to area at p < 10−5, defined as 100%). Compared with the activation areas at p < 10−5, areas measured at p < 5 × 10−5 increased 67.5% for D1 and 81.7% for D3; areas measured at p < 10−4 sizes increased 73.5% for D1 and 135.9% for D3. Scale bars, 1 mm.

Figure 4.

Optical imaging of area 3b and area 1 in response to digit stimulation in the same squirrel monkeys imaged with fMRI. a–c, e, f, Focal activations were obtained in single condition optical images in response to D4 (a), D3 (b), and D2 (c) stimulation in monkey 1, and to D2 (e) and D1 (f) stimulation in monkey 2. Stronger activations in area 3b and weaker ones in area 1 are seen in monkey 1 (a–c). The approximate border between areas 3b and 1 is indicated by the dotted line in d and g. Small color dots (a–g) indicate the electrode penetration locations for electrophysiological recording. Digits recorded are indicated by color code: D4, blue; D3, green; D2, red; and D1, orange. Summary overlays of a–c in monkey 1 and e–f in monkey 2 are shown in d and g, respectively. Scale bars, 1 mm. m, Medial; p, posterior.

Figure 5.

Effect of thresholding level on the location and size of activation revealed by optical imaging. a, d, Raw single-condition images of activations in response to D3 (a) and D4 (d) stimulation (sum of 53 trials) are clipped at two SDs and displayed. b, e, Activation areas are defined by outlining pixels for which the signal change was significant at p < 10−5(red), p < 10−4(green), and p < 10−3(blue) for D3 (b) and D4 (e), respectively. c, f, Activations derived at various thresholds are overlaid for comparison of activation size. Activation area as a function of threshold p values (normalized to the activation area at p < 10−5 = 100%) is shown in g. Scale bar, 1 mm.

Figure 6.

Comparison of somatotopic maps of fingerpads generated by fMRI, intrinsic optical imaging, and electrophysiology in the same animal. a–e, fMRI and optical maps were coregistered using landmarks (e.g., central and lateral sulci) visible on the cortical surface and identifiable vascular markers (black arrows) in both the T2*-weighted structural MR (a, b, d) and optical blood vessel images (c, e). g, h, j, k, Comparison of the composite activation maps obtained from fMRI (g, j) and optical imaging (h, k) for digits D2 (red), D3 (green), and D4 (blue) in monkey 1, and for D2 (red) and D1 (orange) in monkey 2, respectively. b–e are generated from monkey 1. Somatotopic maps of digits obtained from fMRI and optical imaging were overlaid in f (monkey 1) and i (monkey 2) for comparison. Green dots (h, k) indicate the electrode penetration locations where corresponding digit representation were identified. Scale bars: a, 5 mm; b–k, 1 mm. In monkey 1 (g), activations of D4 (blue) and D3 (green) are thresholded at p < 10−7 and D2 (red) at p < 10−8. j, In monkey 2, activations of D1 (orange) and D2 (red) are thresholded at p < 10−8.

Figure 7.

3D plots of fMRI and OIS activations. a–h, Four pairs of statistical activation maps of fMRI (top) and OIS (bottom) runs are shown: D3 and D4 from animal 1 (a–d) and D1 and D2 from animal 2 (e–h). Each pair of fMRI and OIS maps are aligned and displayed with the same field of view. The x- and y-axis indicate the aligned imaging plane in mm scale, and the _z_-axis indicates the t value associated with the mean difference between stimulus and baseline conditions in both fMRI and OIS maps. The mesh is color coded to reflect t value (blue, t ≤ 0; yellow/red, t > 0). For fMRI maps, corresponding p values were t = 5, p = 5 × 10−7; t = 8, p = 10−14; t = 10, p = 10−20; and t = 15, p = 10−35, respectively. For optical images, corresponding p values were t = 2, p = 0.02; t = 3, p = 10−3; and t = 4, p = 10−4, respectively. Arrows indicate activation center locations. Coordinates corresponding to t value peaks were used to calculate offsets between locations of fMRI and OIS activations.

Figure 8.

fMRI reveals the representation of funneling illusory stimuli in area 3b. Two cases are shown (case 1, left; case 2, right). a, b, e, f, For each case, stimulation of single digits D3 (a, b) and D4 (e, f) generated single focal activations organized somatotopically in area 3b (a and e within the blue outline shown in a, shown expanded in b, f). c, d, Stimulation of paired adjacent digits (D3+D4) generated a single focal activation in area 3b, located between the sites of activation during stimulation of D3 (b) and D4 alone (f) (corresponding activation centers are indicated by red dashed lines). Activation is also evident in area 1 (c, d) during simultaneous stimulation of D3+D4 in this case. The color scale from red to yellow (case 1, p < 10−7 to p < 10−15; case 2, p < 10−10 to 10−13; see respective color bar scales) in each activation map indicates the significance level. Scale bars, 1 mm. p, Posterior; l, lateral.

Figure 9.

Comparison of amplitudes of regional fMRI signal change during single and simultaneous stimulation of adjacent (a–c) and nonadjacent (d–f) digits. Spatially shifted (centered) activation was observed in area 3b (a) during simultaneous stimulation of adjacent digits (D1+D2) whereas two separate activations are obtained at each site (D2 and D4) during simultaneous stimulation of nonadjacent digits (D2+D4) (d). Quantitative analysis of the percentage signal change at sites corresponding to individual activation of D1 alone, D2 alone, and D1+D2 (a), or D2 alone, D4 alone, and D2+D4 (d) are shown in b and c (adjacent digit pair) and e and f (nonadjacent pair), respectively. The strongest percentage BOLD signal changes were obtained at D1 or D2 activation sites during individual stimulation of adjacent D1 (b, white columns) or D2 (b, dark columns), and at D2 or D4 activation sites during individual stimulation of nonadjacent D2 (e, white columns) or D4 (e, dark columns). In contrast, the strongest percentage BOLD signal change (c, middle black column) is observed at the midway location (regions analyzed are shown as white boxes in a and d) during simultaneous stimulation of adjacent digits D1+D2 whereas comparable percentage signal changes (e, f) were obtained at D2 and D4 locations (d) during simultaneous stimulation of nonadjacent digits D2+D4. The color scale from red to yellow in each activation map indicates the significance level from p = 10−7 to 10−15 in both a and d. Scale bar, 1 mm.

References

1. Cannestra AF, Pouratian N, Bookheimer SY, Martin NA, Beckerand DP, Toga AW. Temporal spatial differences observed by functional MRI and human intraoperative optical imaging. Cereb Cortex. 2001;11:773–782. - PubMed
1. Cannestra AF, Pouratian N, Forage J, Bookheimer SY, Martin NA, Toga AW. Functional magnetic resonance imaging and optical imaging for dominant-hemisphere perisylvian arteriovenous malformations. Neurosurgery. 2004;55:804–812. - PubMed
1. Chen LM, Friedman RM, Ramsden BM, LaMotte RH, Roe AW. Fine-scale organization of SI (area 3b) in the squirrel monkey revealed with intrinsic optical imaging. J Neurophysiol. 2001;86:3011–3029. - PubMed
1. Chen LM, Heider B, Williams GV, Healy FL, Ramsden BM, Roe AW. A chamber and artificial dura method for long-term optical imaging in the monkey. J Neurosci Methods. 2002;113:41–49. - PubMed
1. Chen LM, Friedman RM, Roe AW. Optical imaging of a tactile illusion in area 3b of the primary somatosensory cortex. Science. 2003;302:881–885. - PubMed

High-resolution maps of real and illusory tactile activation in primary somatosensory cortex in individual monkeys with functional magnetic resonance imaging and optical imaging - PubMed (original) (raw)