TY - JOUR
AU1 - Hu, Jia, Ming
AU2 - Qian, Mei, Zhen
AU3 - Tanigawa,, Hisashi
AU4 - Song, Xue, Mei
AU5 - Roe, Anna, Wang
AB - Abstract Traditional electrical stimulation of brain tissue typically affects relatively large volumes of tissue spanning multiple millimeters. This low spatial resolution stimulation results in nonspecific functional effects. In addition, a primary shortcoming of these designs was the failure to take advantage of inherent functional organization in the cerebral cortex. Here, we describe a new method to electrically stimulate the brain which achieves selective targeting of single feature-specific domains in visual cortex. We provide evidence that this paradigm achieves mesoscale, functional network-specificity, and intensity dependence in a way that mimics visual stimulation. Application of this approach to known feature domains (such as color, orientation, motion, and depth) in visual cortex may lead to important functional improvements in the specificity and sophistication of brain stimulation methods and has implications for visual cortical prosthetic design. brain–machine interface, electrical microstimulation, optical imaging, orientation map, visual cortex Introduction For decades, electrical microstimulation has been widely used in neuroscience research to study brain function and behavior (Bosking et al. 2017). From the initial groundbreaking mapping studies by Penfiled and Perot (1963) to studies demonstrating the ability to modulate behavior in functionally specific manners (Salzman et al. 1990; Romo et al. 1998), electrical stimulation has held a central role in mapping and modulating brain function. Its development for deep brain stimulation in humans has revolutionized treatment of diseases such as Parkinson’s and has solidified its place in the clinical toolbox (Wichmann and Delong 2006; Lewis and Rosenfeld 2016). Although its utility for achieving neuronal modulation is uncontested, traditional electrical stimulation methods have not taken advantage of known fundamental organizations in the brain. In many large mammals such as humans, nonhuman primates, and cats, sensory representation is based on submillimeter functional units in the cerebral cortex that encode feature-specific information. In the visual system, these include features such as contour orientation, surface color, visual depth, and motion direction (Chen et al. 2008; Lu and Roe 2008; Lu et al. 2010; Tanigawa et al. 2010; Li et al. 2013; Hu et al. 2018). Such feature-specific domains have specific parallel circuitries in early visual areas (Roe et al. 2012). Object representation therefore arises from visual activation of multiple networks of feature-specific domains. Our hypothesis is that the activation of such feature-specific networks can also be achieved by nonvisual stimulation methods, such as focal electrical stimulation. Here, we address this hypothesis by focusing on visual orientation processing networks (Chen et al. 2016; Fang et al. 2019). It is well-known that in visual cortex orientation domains of similar orientation selectivity form orientation-selective networks. Specifically, orientation domains in the visual cortices of cats and primates contain clusters of neurons with similar orientation preference which are anatomically linked in patchy networks via local horizontal connections (Fig. 1, Gilbert and Wiesel 1989; Levitt et al. 1994; Sincich and Blasdel 2001; Stettler et al. 2002). Given this connectivity, we expect that stimulation of one orientation domain (Fig. 1, yellow lightning) will lead to activation of other orientation domains of matching selectivity (via red arrows). Moreover, given the push–pull relationship of matching and nonmatching orientation networks (Fig. 1, gray ovals, Morrone et al. 1982; DeAngelis et al. 1992), one might expect different effects on nonmatching orientation domains. Domains of unrelated functionality (Fig. 1, blue ovals) would not be affected. Given the submillimeter scale of each domain, electrical stimulation that is not focal will, via current spread, activate multiple functional domains of different selectivity and thus fail to evoke feature-specific activation. Figure 1 Open in new tabDownload slide Predicted outcome of focal, targeted electrical stimulation. Visual cortex is composed of arrays of feature-specific domains (columns), which are submillimeter in size (human ~ 800 μm, cat ~ 800 μm). Three such arrays are depicted by orange (e.g., horizontal orientation network), gray (e.g., vertical orientation network), and blue (e.g., color network in primates) ovals. Due to existing selective horizontal connections (red arrows), electrical stimulation (yellow lighting) of a single column is predicted to activate columns of similar functionality (orange) and have differential effects on those of opposing selectivity (gray). Other columns (blue array) that are not functionally related to the stimulation site will not be strongly modulated. Electrical stimulation that is not focal (i.e., not confined to a single domain) will activate multiple functional networks of different selectivities and thus fail to evoke feature specific activation (Murasugi et al. 1993). Figure 1 Open in new tabDownload slide Predicted outcome of focal, targeted electrical stimulation. Visual cortex is composed of arrays of feature-specific domains (columns), which are submillimeter in size (human ~ 800 μm, cat ~ 800 μm). Three such arrays are depicted by orange (e.g., horizontal orientation network), gray (e.g., vertical orientation network), and blue (e.g., color network in primates) ovals. Due to existing selective horizontal connections (red arrows), electrical stimulation (yellow lighting) of a single column is predicted to activate columns of similar functionality (orange) and have differential effects on those of opposing selectivity (gray). Other columns (blue array) that are not functionally related to the stimulation site will not be strongly modulated. Electrical stimulation that is not focal (i.e., not confined to a single domain) will activate multiple functional networks of different selectivities and thus fail to evoke feature specific activation (Murasugi et al. 1993). Based on this known cortical organization, we aimed to design a technology that selectively activates these circuits to achieve feature-specific activation. We developed a method that fulfills several criteria: (1) activation that is reliable across multiple stimulations, (2) can be quantitatively controlled in an intensity-dependent fashion, (3) is functionally selective, that is, achieves visual activation patterns similar to which obtained with normal visual stimulation, and (4) is sufficiently focal that it is confined to single functional domains. These results are the first to demonstrate the possibility of using focal and targeted electrical stimulation in visual cortex to mimic the highly specific patterns of cortical response evoked by visual stimulation. Materials and Methods Animal Procedures All procedures were in compliance with and approved by the Institutional Animal Care and Use Committee of Zhejiang University, and followed the guidelines of the National Institute of Health Guide for the Care and Use of Laboratory Animals. Experiments were performed in four hemispheres of three male cats (weighing 2–4 kg). Anesthesia in cats was induced with ketamine (10 mg/kg) and maintained with isoflurane (1.5–2%). Animals’ vital signs were monitored and temperature was maintained during the surgery and experiment. A craniotomy (Horsley-Clarke coordinates A10-P5, L0-L15) and durotomy were performed to expose visual cortical areas 17 and 18 (Wang et al. 2011; Li et al. 2017; Fig. 1A). A piece of cover glass, with a hole for insertion of an electrode, was placed on top of the cortex to suppress physiological motion (Fig. 1B and C). After the stabilization of cortex, cats were anesthetized with propofol (induction 5–10 mg/kg, maintenance 5–10 mg/kg/h, i.v.) and were paralyzed (vecuronium bromide, induction 0.25 mg/kg, maintenance 0.05–0.1 mg/kg/h, i.v.). The pupils were dilated with atropine sulfate (1%) and the nictitating membranes were retracted with neosynephrine (5%). The eyes were corrected with appropriate contact lenses to focus on a monitor 57 cm in front of the animal. Electrical Stimulation We stimulated with glass-coated tungsten electrodes (0.5–1.5 MΩ at 1 kHz) with tip diameters of 20–40 μm. Electrodes were carefully positioned with a manipulator (Narishige, Stereotaxic Micromanipulator SM11) under the optical imaging camera. We used trains of short pulses (5 ms intervals) at 200 Hz, lasting 150 or 200 ms. Electrodes were carefully inserted into the desired location guided by the imaging camera. In our recordings, the angle between the electrode and cortical surface was around 45° (Fig. 2). The electrodes were advanced into the cortex with a distance along the electrode track of 500–700 μm, resulting in a depth of 350–500 μm beneath the cortical surface, ending up about 300–500 μm lateral to the point of entry. The electrodes were thus placed so that the stimulation sites were near the center of the targeted orientation domain (~1 mm in size) and at a depth corresponding to layers 2/3 (Matsubara et al. 1987). Current pulses were provided by a stimulus isolator (WPI, Stimulus Isolator A365) with biphasic form. Figure 2 Open in new tabDownload slide Illustration of experiment design. (A) Current was delivered to the cortex in area 18 through a microelectrode. Lower panel shows the cortical surface of imaging areas. The location of the electrode tip is marked by red star. A, anterior; M, medial. (B) A coverglass was used to stabilize the cortex. Electrodes were inserted into the cortex through a hole in the glass. (C) Stimulation and imaging paradigm. In each trial, optical imaging lasted for 4 s. About 0.5 s later, after imaging began, a train of current pulses were delivered to the cortex for 150 or 200 ms. In cases where effects of electrical and visual stimulation were compared, visual stimuli were also presented for 3.5 s. D. Magnified view of the region around tip site marked by dashed red square in (A). The lower panel shows the regions in which cortical responses are significantly (two-tailed t-test, P < 0.05) enhanced by electrical stimulation compared with blank condition. Scale bar 1 mm. Figure 2 Open in new tabDownload slide Illustration of experiment design. (A) Current was delivered to the cortex in area 18 through a microelectrode. Lower panel shows the cortical surface of imaging areas. The location of the electrode tip is marked by red star. A, anterior; M, medial. (B) A coverglass was used to stabilize the cortex. Electrodes were inserted into the cortex through a hole in the glass. (C) Stimulation and imaging paradigm. In each trial, optical imaging lasted for 4 s. About 0.5 s later, after imaging began, a train of current pulses were delivered to the cortex for 150 or 200 ms. In cases where effects of electrical and visual stimulation were compared, visual stimuli were also presented for 3.5 s. D. Magnified view of the region around tip site marked by dashed red square in (A). The lower panel shows the regions in which cortical responses are significantly (two-tailed t-test, P < 0.05) enhanced by electrical stimulation compared with blank condition. Scale bar 1 mm. Visual Stimuli Visual stimuli were created with ViSaGe (Cambridge Research Systems Ltd) and presented on a monitor refreshed at 60 Hz and positioned 57 cm from the cat’s eyes. Full screen drifting sinusoidal gratings (0.2 cycle/degree, 5 Hz) were used to obtain orientation maps in area 18. Gratings were presented in a randomly interleaved fashion in 1 of 4 orientations (0°, 45°, 90°, and 135°) moving bidirectionally. Intrinsic Signal Optical imaging Imaging was performed with Imaging system 3001 (Optical Imaging Inc.). Images were obtained from the camera under 630 nm light. Frames were acquired at 4 Hz for 4 s synchronized to respiration. About 0.5 s after imaging began, electrical stimulation was turned on (Fig. 2C). There were three types of stimulation conditions: visual stimulation only, electrical stimulation only, and blank (timing indicated in Fig. 2C). Each condition was presented for 30–50 trials, pseudorandomly interleaved with at least a 7-s intertrial interval. Visual stimulation was presented binocularly. A blood vessel map of the cortex (Fig. 2A) was also acquired under 596 nm light for analysis. Data Analysis Data analysis was performed with Matlab (The Mathworks, Matlab-R2012b). To show the cortical response timecourse, we used “single-frame maps.” For each frame, the gray value of each pixel was calculated first using the following functions: dR/R = (Fx – F0)/F0, where F0 is the average reflectance value of the first two frames (taken before electrical stimulation onset), Fx is the reflectance value corresponding to frame X (X = 1–16). To remove low-frequency noise (e.g., gradations in illumination across the cortex, slow 0.1 Hz oscillations, slow physiological fluctuations), maps were low-pass filtered (Gaussian filter, 3–8 pixel diameter), and low-frequency noise was reduced by convolving the map with a 150–200 pixel diameter (~1.2–1.6 mm) circular filter and subtracted from the original maps. Pixel values were clipped by mean ± 2 standard deviation (SD); the pixel values larger than mean + 2 SD were truncated to mean + 2 SD. Response time courses were obtained from the regions of interest (see Fig. 4). The peak value between frames 5–16 was used as the response amplitude for each trial (see Fig. 4H). To show the response maps induced by stimulation (electrical/visual), we calculated the dR/R values in the conditions with stimulation and the values in the blank conditions. Then with the following formula, |${\Delta \mathrm{R}}_{\mathrm{i}}=\big({R}_{istim}-{R}_{iblank}\big)\times \sqrt{N}/{S}_i$|⁠, we obtained the response changes between the stimulation on and off conditions to establish the modulation effects. Ristim is the dR/R value in the condition with stimulation of pixel i, Riblank is the dR/R value in the condition without stimulation of pixel i, N is the number of trials, Si is the standard deviation of (Ristim − Riblank) and the division is used to reduce the contribution excessively large signals from blood vessels. Maps were also filtered and clipped as described above. For electrical stimulation, maps were calculated from average of the fifth frame to the eighth frame (Fig. 6); for visual stimulation frames 8–16 were used (Fig. 3). Based on our image acquisition frame rate (250 ms/frame), we observed initial activation at 250 ms after electrical stimulation onset (i.e., in the second frame after stimulation onset). We therefore used the size of activation in the second frame after stimulation onset as the evoked activation size of electrical stimulation. That is, this stimulation site contained the significant pixels activated simultaneously at the electrode site (within the limits of our 4 Hz sampling rate). An ellipse was drawn around these pixels, the height and width of this ellipse were used to represent the size of the region that was directly stimulated by electrical stimulation (see Fig. 6B). Figure 3 Open in new tabDownload slide Focal electrical stimulation produces maps similar to visual orientation maps. Each row represents the results from one case. (A–F) Orientation maps from cat area 18 (dark vs. light pixels: A–C, 0° vs. 90°; D–F, 45° vs. 135°). (G-I) Single condition maps of focal electrical stimulation compared with blank (50 μA stimulation). (J-L) The same maps as in (G-I), except filtered with same parameters as orientation maps in (A–F). Yellow star: electrical stimulation site (position of the tip after correcting for the angle of insertion and point of entry, see Methods). Red and blue dots: reference dots to compare domains activated by electrical stimulation and visual orientation domains. Inset next to (J-L): oblique lines in the upper right represent the preferred orientation of each pixel in an area centered on stimulation site (n = 121, 11 × 11 pixels); red line and number below represents the average orientation preference. (M–O) Cortical response tuning curve of electrical stimulation (gray lines). The strongest responses are marked by red dots in each tuning curve (see text). The dR/R% values corresponding to 0–1 normalized responses were M: 0.04–0.07, N: 0.05–0.09, O: 0.05–0.09. Scale bar 1 mm. Figure 3 Open in new tabDownload slide Focal electrical stimulation produces maps similar to visual orientation maps. Each row represents the results from one case. (A–F) Orientation maps from cat area 18 (dark vs. light pixels: A–C, 0° vs. 90°; D–F, 45° vs. 135°). (G-I) Single condition maps of focal electrical stimulation compared with blank (50 μA stimulation). (J-L) The same maps as in (G-I), except filtered with same parameters as orientation maps in (A–F). Yellow star: electrical stimulation site (position of the tip after correcting for the angle of insertion and point of entry, see Methods). Red and blue dots: reference dots to compare domains activated by electrical stimulation and visual orientation domains. Inset next to (J-L): oblique lines in the upper right represent the preferred orientation of each pixel in an area centered on stimulation site (n = 121, 11 × 11 pixels); red line and number below represents the average orientation preference. (M–O) Cortical response tuning curve of electrical stimulation (gray lines). The strongest responses are marked by red dots in each tuning curve (see text). The dR/R% values corresponding to 0–1 normalized responses were M: 0.04–0.07, N: 0.05–0.09, O: 0.05–0.09. Scale bar 1 mm. Masks for the regions that were significantly modulated by the electrical microstimulation were obtained from two-tailed t-tests between the electrical stimulation on and off conditions. Pixels with P < 0.05 were selected. The same thresholds were used for orientation maps (e.g., Fig. 5B and C). Cortical response profiles were drawn in the way same as described in previous studies (cf. Lu et al. 2010). In brief, we used the 4 orientation maps to estimate the orientation preference of each pixel and classified these pixels in 18 orientation bins, resulting in 0–180° orientation vector maps. The orientation angle maps (Fig. S2) were obtained first through a vector analysis (Bosking et al. 1997); then a cortical response profile for a given map was obtained by calculating the pixel values corresponding to different orientation domains (18 orientation bins from 0 to 170°, each iso-orientation domain covers a 30° wide range, i.e., vertical domain ranges from 75° to 105°, Fig. S2). The cortical responses were passed to a least-square nonlinear regression function (“nlinfit” in Matlab, Mathworks) that fits the data with the following formula. |$y=a+b\times{e}^{c\times \big[\cos\;2\big(x-d\big)-1\big]}$|⁠; x is the orientation, y is the corresponding cortical response, a is the baseline offset, b, c, d determine the amplitude, shape, and position of the curve, respectively (see Fig. 3, M–O). All the goodness values (R2) were > 0.7. Statistics Matlab (The Mathworks, Matlab-R2012b) was used for statistical analyses. For picking the pixels that are significantly modulated by electrical stimulation, we performed two-tailed t-tests (P < 0.05) between the two comparison groups (stimulation on and off). Results Determining Electrical Stimulation Parameters for Targeted Activation To visualize the effects of electrical stimulation in anesthetized cat visual cortex (area 18), we prepared an optical window through which we could insert a microelectrode (Fig. 2A and B). Based on previous studies of the influence of electrical stimulation on behavior (Salzman et al. 1990; Murasugi et al. 1993; Celebrini and Newsome 1995; Britten and Wezel 1998; DeAngelis et al. 1998; Murphey and Maunsell 2007), we selected a range of stimulation intensities (25–100 μA), pulse widths (200, 500 μs, constant-current biphasic square pulses), frequency (200 Hz), stimulation durations (150, 200 ms), and tested a matrix of parameters (Fig. 2C, an example of our stimulation parameters). We targeted superficial cortical layers (depth 350–500 um) which are known to contain orientation specific horizontal networks (Matsubara et al. 1987; Gilbert and Wiesel 1989; Bosking et al. 1997). To avoid excessive current spread (activating multiple columns) caused by higher current amplitude (Murasugi et al. 1993), we limited our stimulation current amplitudes below 100 μA. Then, using intrinsic signal optical imaging, we obtained images of the response evoked at the electrode tip (Fig. 2D). We found that, at the electrode tip, the electrical stimulation produced a focal activation approximately 1 mm in size (Fig. 2D). This demonstrated that, with suitable parameters, electrical stimulation could be confined to single submillimeter-sized cortical columns. Based on these results, we adopted 25, 50, 75, and 100 μA to evoke cortical responses in most of our experiments. Relationships Between Visual and Electrical Stimulation Maps Visual Maps To illustrate the relationship between visual and electrical stimulation maps, we first obtained maps of cortical response to normal visual stimulation. As shown in Figure 3A–F, visual cortical maps were obtained in response to drifting gratings presented at four orientations (three cases, A–C: 0° vs. 90°, D–F: 45° vs. 135°). As shown in Figure S1A and B, nonfiltered single condition maps each produce activations (negative reflectance change) at orientation-selective sites; subtraction of orthogonal conditions produces an orientation map (Fig. S1C). Following standard methods (e.g., Bonhoeffer and Grinvald 1996; Lu et al. 2010; Tanigawa et al. 2010; Li et al. 2013), images were then spatially filtered to remove low-frequency noise (i.e., images were convolved with a ~1.2–1.6 mm circular filter and subtracted from the original maps); this serves to normalize the image to the local response differences within the image and make differential response more evident (see Methods, compare Fig. S1C and D). Electrical Stimulation Maps Using these orientation maps as a reference for cortical responses to electrical stimulation, we then targeted a microelectrode to a single orientation domain (Fig. 3G, 45°; Fig. 3H, 0°; Fig. 3I, ~20°; note that the precise targeted orientation was determined subsequently using procedures described below). Figure 3G–I (yellow stars) show the results of this electrical stimulation (in the absence of visual stimulation, single condition maps). Images were spatially filtered in the same manner as visually evoked orientation maps, to remove low-frequency noise (Fig. 3J–L). The precise orientation selectivity of the targeted site was quantified by averaging the preferred selectivity of 11 × 11 pixels (~8–10 μm/pixel) around the tip site (insets at right of J, K, L; Case 1: 43 ± 12°, Case 2: 4 ± 2°, Case 3: 26 ± 16°, mean ± SD, see Methods). We predicted that domains of matching preference would be activated. To permit direct alignment of response patterns, orientation maps and electrical stimulation patterns were obtained within the same experiment session. Comparison of Maps Following electrical stimulation of single orientation domains, we observed that there were activations away from the stimulated site (unfiltered, Fig. 3G–I). To examine whether the response patterns evoked by electrical stimulation were similar to the orientation maps evoked by the matching orientation grating, we conducted the same spatial filtering as that conducted on the orientation maps (Fig. 3J–L). This revealed that locations with significantly activated pixels tended to overlie orientation-matched locations (see red dots in Fig. 3, compare D and J, B and K, F and L). For example, when the stimulation was conducted in an oblique (45°) domain (Fig. 3J), the pattern was similar in appearance to the 45° versus 135° orientation map (Fig. 3D). When a horizontal (0°) domain was stimulated (Fig. 3K), the response pattern shared a similar distribution as the 0° versus 90° orientation map (Fig. 3B). Quantification of Orientation Selectivity of Population Response: Method This impression is further supported by quantitative assessment of the cortical population response (Lu et al. 2010). For each of these three cases, we used the 4 orientation maps to estimate the orientation preference of each pixel and classified these pixels in 18 orientation bins, resulting in 0–180° orientation vector maps (a standard method of inferring complete orientation preference maps, where each pixel is color-coded by the vector sum of response to each of the four orientations, Bosking et al. 1997; Lu et al. 2018) (Fig. S2A–C). We then identified orientation domains at each of 0, 10, 20…170° preferences (by selecting all pixels corresponding to a single orientation ±15°; note that averaged tuning width of orientation neurons in Area 18 is approximately 30°, Hammond and Andrews 1978). As shown in Figure S2D, for each case, the population tuning curve (which plots the average response of all pixels at each 10° bin) showed peak population response at the expected orientations (e.g., 0° grating, orange line in Fig. S2D, produces strongest activation at 0° bin). We then applied this procedure to the electrical stimulation maps and produced a population tuning curve of electrical stimulation response (graphs in Fig. 3M–O). As shown in Figure 3M, electrical stimulation produced maximal activation at a subset of the pixels, which peaked at the 40° orientation bin (peak preference at 41° when fit with a curve, gray line), very close to the orientation preference of the stimulation site (43°). A similar result was obtained for the other two cases (Fig. 3N: stimulation of a 4° domain produced a quantitative peak at the 170° bin, peak at 7° when fit with curve; Figure 3O: stimulation of a 26° domain produced a quantitative peak at 10° bin, peak at 12° when fit with curve). Thus, our results suggest that the electrically evoked spatial response patterns strongly depended on the stimulation site location. Thus, consistent with our prediction, the effect of electrical stimulation was largely confined to domains with matching orientation selectivity. This approach also demonstrates the ability to decode the stimulated pattern. That is, the tuning curves (Fig. 3M–O) generated from the stimulated images can be used to decode which orientation is being processed in the cortex, something key to using a neuroprosthetic in an informative and accurate fashion. Intensity-Dependent Responses at Stimulated and Connected Sites Because visual stimulus features vary in intensity (e.g., brightness, speed, and color saturation), neural population responses in cortical functional domains encode this accordingly in an intensity-dependent fashion (Lu and Roe 2007). For a neural prosthetic to achieve visual capabilities effectively, it should be able to modulate neural response by varying stimulation intensity, leading to increasing response with increasing stimulation intensity. We also expect that increases in induced response at the stimulated site would induce increasing response at connected sites, thereby achieving intensity-dependent response of the functionally selective network. To examine this, we applied increasing stimulation intensities to single orientation domains and measured the optical response amplitude at the stimulation site and connected sites. Original unfiltered images are shown in Figure S3 and filtered images, which enhance visualization of the differences, are shown in Figure 4. As shown in Figure 4A, D, and G, increasing stimulation intensity resulted in spatially similar activation maps. In Case 1 (Fig. 4A), stimulation (yellow star) at 25, 50, and 75 μA results in focal activation at the same sites (sites 2– 4). In Case 2 (Fig. 4D), stimulation (yellow star) at 25 μA did not achieve detectable response, but stimulation at 50, 75, and 100 μA resulted in focal activation at the same sites (sites 2 and 3). In Case 3 (Fig. 4G), stimulation (yellow star) at 50 and 75 μA also resulted in focal activation at the same sites (sites 1–6). This demonstrates that activation of a spatially stable network can be evoked by different intensities of single site stimulation. Figure 4 Open in new tabDownload slide Intensity dependence. Each row illustrates results from one case. (A, D, G) Cortical response patterns in response to different current amplitudes (marked at top of each panel, 0 μA: cortical response in blank condition). Forth panel of G: single condition map (without filtering procedures). Yellow stars: stimulation sites. Red dots: activated sites. Electrode is outlined by red lines. Scale bar 1 mm. (B, E, H) Response timecourses of stimulation sites. (C, F, I) Response timecourses of connected sites. All timecourses are derived from nonfiltered maps. Different gray and color lines: different current amplitudes. Error bar: standard error of the mean (SEM). Amplitude: peak of the initial dip (e.g., current amplitude: 50 μA, red circle, at ~2 s). Error bar: SEM. Case 3 in Figure 4 and Case 2 in Figure 3 are from the same case. (J-M) Similarity of visual and electrical stimulation amplitudes. (J) Visual orientation map (horizontal minus vertical). (K) Colored pixels: visual horizontal preferring pixels (sites 1–5, two-tailed t-test horizontal vs. vertical, P < 0.05). (L) Colored pixels: electrically evoked pixels from stimulation at site 1 (sites 1–5, two-tailed t-test, stimulation vs. blank, P < 0.05). (M) Plot of amplitudes of visual (black dots) and electrical (red dots: 50 μA, green dots: 75 μA) stimulation at site 1 (stimulation site) and each of sites 2–5 (connected sites). Figure 4 Open in new tabDownload slide Intensity dependence. Each row illustrates results from one case. (A, D, G) Cortical response patterns in response to different current amplitudes (marked at top of each panel, 0 μA: cortical response in blank condition). Forth panel of G: single condition map (without filtering procedures). Yellow stars: stimulation sites. Red dots: activated sites. Electrode is outlined by red lines. Scale bar 1 mm. (B, E, H) Response timecourses of stimulation sites. (C, F, I) Response timecourses of connected sites. All timecourses are derived from nonfiltered maps. Different gray and color lines: different current amplitudes. Error bar: standard error of the mean (SEM). Amplitude: peak of the initial dip (e.g., current amplitude: 50 μA, red circle, at ~2 s). Error bar: SEM. Case 3 in Figure 4 and Case 2 in Figure 3 are from the same case. (J-M) Similarity of visual and electrical stimulation amplitudes. (J) Visual orientation map (horizontal minus vertical). (K) Colored pixels: visual horizontal preferring pixels (sites 1–5, two-tailed t-test horizontal vs. vertical, P < 0.05). (L) Colored pixels: electrically evoked pixels from stimulation at site 1 (sites 1–5, two-tailed t-test, stimulation vs. blank, P < 0.05). (M) Plot of amplitudes of visual (black dots) and electrical (red dots: 50 μA, green dots: 75 μA) stimulation at site 1 (stimulation site) and each of sites 2–5 (connected sites). We further analyzed the amplitudes of response at each of these sites. As predicted, the amplitude of peak response increased with increasing electrical stimulation intensity at the stimulation site (yellow stars, see Fig. 4B and E timecourses, light gray, gray, dark gray lines). Moreover, at connected sites, as shown in Figure 4C and F, increasing stimulation intensity led to increasing response amplitude at connected sites (C: sites 2–4, F: sites 2–3, light gray, gray, dark gray, and black lines). In Case 3, (Fig. 4G), consistent with other cases, we observed increasing response in the stimulated domain (Fig. 4G, site 1, yellow stars; Fig. 4H, light and dark gray lines) as well as in connected sites (Fig. 4G, sites 2–6; Fig. 4I, peak amplitudes of 50 μA: 0.07% and of 75 μA: 0.09%). Thus, the stronger the electrical stimulation, the greater the signal amplitude at both stimulated and at orientation-matched connected sites. Amplitude Similarity of Visual and Electrical Stimulation We also noted the similarity of response amplitude of electrical and visual stimulation (Fig. 4J). As shown in the response timecourses, the optical signal evoked by electrical stimulation was similar in temporal profile to normal visual cortical signal (the “initial dip”), exhibiting amplitudes of 0.01–0.2% and timecourses that peaked 1–2 s poststimulus onset (e.g., Grinvald et al. 1986, 2000; Roe 2007; Lu et al. 2017). The similarity of optical signals to electrical and sensory stimulation has previously been noted (Chen et al. 2008) and suggested that electrical stimulation can achieve natural vision-like effects in the brain. Previous studies, however, did not directly compare responses between electrical and visual stimulation, and so were unable to determine whether electrical stimulation can induce response amplitudes similar to visual stimulation. To address this question, we compared the significant pixels evoked by visual stimulation (Fig. 4K) with that evoked by electrical stimulation (Fig. 4L, at 50 μA). Each of the horizontal domains (Fig. 4K, sites 1–5) were matched with the electrically activated domains (Fig. 4L, sites 1–5). As shown in Figure 4M, excluding the electrically stimulated domain (site 1), for each of the domains (sites 2–5), electrical stimulation (red dots, 50 μA; green dots, 75 μA) evoked comparable (on the order of 0.1%) response amplitudes as visual stimulation (black dots). This shows that, at identical sites in visual cortex, electrical stimulation can evoke similar response amplitude as visual stimulation. These initial characterizations suggest that it is feasible to use electrical stimulation in a focal, domain-selective, and intensity-dependent manner. Weak Enhancement of Opposing Orientation Networks We found that focal electrical stimulation could also affect orthogonal domains. We highlight Case 2 (same case as in Fig. 4G). As we previously described in Figure 4I, when a horizontal domain is stimulated (yellow star), the strongly activated pixels (dR/R ~ 0.1%) tend to overlap with orientation domains of matching selectivity (Fig. 5B, red, horizontal domains outlined in white). In addition, this stimulation led to weak activation of a second set of pixels (Fig. 5C, left, dR/R ~ 0.05%). The complementary spatial relationship of these two sets of pixels was further confirmed by comparing the distributions of orientation difference between the stimulation site and the pixels modulated by electrical stimulation (Fig. 5D, red bars: matched, blue bars: nonmatched). For matched pixels, the distribution peak was centered around 0°, while for nonmatched, the distribution peak was centered around 90°. Figure 5 Open in new tabDownload slide Orientation specific activation and suppression. (A) Visual orientation map. (B and C) Overlay of orientation domains and electrically stimulated domains. Pixels identified based on filtered maps. (B) White outlines: horizontal orientation domains. Red pixels: significantly enhanced pixels (relative to the filtered mean) following electrical stimulation (yellow star) of a horizontal domain (P < 0.05, two-tailed t-test). (C) White outlines: vertical orientation domains. Blue pixels: weakly enhanced pixels (relative to the filtered mean) following electrical stimulation of a horizontal domain (P < 0.05, two-tailed t-test). (D) The distribution of preferred orientation difference between stimulation site and the pixels that are significantly modulated by electrical stimulation in (B) and (C) (red bars: strong enhancement, blue bars, weak enhancement). (E) Comparison of timecourses to lower (50 μA) and higher (75 μA) stimulation intensities in single condition maps (nonfiltered) at two locations (sites 7 and 8). (F) The timecourses of both matched (sites 2–6) and orthogonal domains (sites 7 and 8) at two stimulation intensities. Scale bar 1 mm. Figure 5 Open in new tabDownload slide Orientation specific activation and suppression. (A) Visual orientation map. (B and C) Overlay of orientation domains and electrically stimulated domains. Pixels identified based on filtered maps. (B) White outlines: horizontal orientation domains. Red pixels: significantly enhanced pixels (relative to the filtered mean) following electrical stimulation (yellow star) of a horizontal domain (P < 0.05, two-tailed t-test). (C) White outlines: vertical orientation domains. Blue pixels: weakly enhanced pixels (relative to the filtered mean) following electrical stimulation of a horizontal domain (P < 0.05, two-tailed t-test). (D) The distribution of preferred orientation difference between stimulation site and the pixels that are significantly modulated by electrical stimulation in (B) and (C) (red bars: strong enhancement, blue bars, weak enhancement). (E) Comparison of timecourses to lower (50 μA) and higher (75 μA) stimulation intensities in single condition maps (nonfiltered) at two locations (sites 7 and 8). (F) The timecourses of both matched (sites 2–6) and orthogonal domains (sites 7 and 8) at two stimulation intensities. Scale bar 1 mm. Figure 5E and F show that the response timecourses overlying the orthogonal domains (sites 7 and 8) were significantly weaker (peak amplitude, mean ± SD = 0.044 ± 0.006%, sites 7–8 in both 50 and 75 μA) than those over matching domains (red, peak amplitude, mean ± SD = 0.087 ± 0.039%, sites 2–6 in both 50 and 75 μA) (sites 2–6 compared with sites 7–8 in both 50 and 75 μA, Wilcoxon test, P = 0.008). Although weaker, the nonmatched domains were also affected by intensity, such that increasing stimulation intensity resulted in a slightly (Wilcoxon test, P = 0.014, comparing the dR/R values from frame 4 to 12 of sites 7 and 8 in 50 and 75 μA, see Fig. 5E) reduced response amplitude (Fig. 5E peak amplitudes, site 7: 50 μA: 0.051%, 75 μA: 0.041%; site 8: 50 μA: 0.047%, 75 μA: 0.037%), suggesting the presence of a weak suppressive effect. Comparison of these simultaneous effects on red and blue pixels in Figure 5F shows that 50 μA stimulation produces a response differential of approximately 0.03% (cyan vs. orange), while 75 μA stimulation results in a greater approximately 0.05% amplitude difference (blue vs. red) (Wilcoxon test, P = 0.003, comparing the dR/R value differences between sites 2–6 and sites 7–8 from frame 4 to 12 in 50 and 75 μA, see Fig. 5F). Thus, the resulting effect of increasing electrical stimulation leads to an enhancement of response contrast between orthogonal domain networks. Spatiotemporal Development of Activation Pattern and Single Domain Activation Confirmation To further illustrate how electrical stimulation affects the cortical response through time, we illustrate the evolution of the response pattern following electrical stimulation. To make clear the stability of the maps, we illustrate filtered images; nonfiltered images can be seen in Figure S4. As shown in Figure 6, the first appearance of significant response to electrical stimulation appears 0.25–0.5 s (Fig. 6A, F4) after stimulation onset. Following initial focal activation, the map emerges around 0.5 s after stimulation onset, and is maintained over the next second or so, followed by gradual weakening over time, after which, 2 s later (F12–F13), the pattern becomes more noisy (F14–F16), perhaps due to vascular influx of oxygenated blood. This suggests that electrical stimulation can evoke stable, long-lasting (~2 s) representation, consistent with normal visually evoked hemodynamic patterns in the cortex. Figure 6 Open in new tabDownload slide Timecourse and spread of the activation. The frames (F1–F16, sampling rate 4 Hz) recorded with optical imaging showed optical reflectance changes in area 18 during electrical stimulation. Electrical stimulation was given in the third frame (F3). The time interval between two adjacent frames was 0.25 s. Scale bar 1 mm. (B) The response induced at the stimulation site is limited within a relative small region. In this case, the width of the directly affected region is 0.8 mm (blue line), the length is 1 mm (red line). (C) On average, the length of the directly affected region is 0.85 mm, the width is 0.63 mm. Figure 6 Open in new tabDownload slide Timecourse and spread of the activation. The frames (F1–F16, sampling rate 4 Hz) recorded with optical imaging showed optical reflectance changes in area 18 during electrical stimulation. Electrical stimulation was given in the third frame (F3). The time interval between two adjacent frames was 0.25 s. Scale bar 1 mm. (B) The response induced at the stimulation site is limited within a relative small region. In this case, the width of the directly affected region is 0.8 mm (blue line), the length is 1 mm (red line). (C) On average, the length of the directly affected region is 0.85 mm, the width is 0.63 mm. Note that to evoke an orientation-specific response, we used a current intensity to achieve activation within one single orientation domain. We observed that the early activation is focal and appeared similar to the dimensions of the orientation domain. The sizes of the electrically stimulated sites (averaged from 17 stimulation experiments across four hemispheres on three cats in which current amplitude ranged from 50 to 100 μA) measured 0.85 ± 0.25 mm in length by 0.63 ± 0.21 mm in width (mean ± SD) (see Fig. 6, B and C). These sizes are consistent with that of single orientation domains in area 18 (Swindale et al. 1987; Shmuel and Grinvald 1996; Wang et al. 2011), and confirm that, at least initially, the effects of electrical stimulation is confined to a single orientation domain. We also examined the spatial extent of stimulation effect by measuring distances between the stimulation site and activated pixels (see Fig. S5). The histogram of all radially measured distances shows that connected pixels extend up to approximately 3 mm away from the stimulation site. This corresponds to about 3–5° of visual space, consistent with integration fields of horizontal connections within area 18 demonstrated anatomically (Matsubara et al. 1987) and electrophysiologically (Tusa et al. 1979; Albus and Rechmann 1980). Moreover, these histograms show several peaks, rather than 1 peak, reflecting the domain-based aspect of the orientation maps (Matsubara et al. 1987; Gilbert and Wiesel 1989). Discussion Summary We demonstrate that focal electrical microstimulation in vivo can evoke selective activation of cortical networks. Key to this success is (1) the imaging and targeting of orientation domains using intrinsic signal optical imaging and (2) the technical development of focal domain-specific electrical stimulation methodology. We show that cortical responses to electrical stimulation are domain specific, intensity dependent, and exhibit differential effects on opposing orientation networks. Observed intensity dependence included: (1) a robust response at orientation-matched activation sites and (2) a weak response reduction at sites of orthogonal orientation preference. These two effects, when considered together, reveal that increasing electrical stimulation intensity produces greater response contrast between orientation-matched and nonmatched sites. Similar to visual stimulation, activation is reliable and can be similar in response amplitudes. Our study suggests that targeted, focal electrical stimulation is capable of mimicking the complex cortical activation patterns evoked during natural vision. Although a single electrode can only affect limited visual fields, an electrode array with multiple tips would evoke a larger portion of the visual map; such arrays have been designed and applied to record from large areas of cortical hemispheres in human (Ghovanloo and Najafi 2004, 2007; Hochberg et al. 2012). Developing a Novel Paradigm Previous paradigms for electrical stimulation in the cortex have employed a wide range of stimulation parameters (intensities, durations, pulse frequencies, and electrode types) (reviewed in Ranck 1975; Murasugi et al. 1993; Weliky et al. 1995; Tehovnik 1996; Cohen and Newsome 2004; Tehovnik et al. 2004, 2006; Histed et al. 2009; Bosking et al. 2017). These studies have shown that focal electrical stimulation can evoke activation of global networks in visual (Tolias et al. 2005; Moeller et al. 2017) and somatomotor cortex (Stepniewska et al. 2011). Sensory discrimination performance can also be systematically influenced, such as shifting the perceived direction of motion in an ambiguous moving random dot pattern (Salzman et al. 1990; Murasugi et al. 1993; Celebrini and Newsome 1995; Britten and Wezel 1998; DeAngelis et al. 1998; Murphey an Maunsell 2007). However, most of the studies focused on describing modulation effects of electrical stimulation when added to ongoing visual stimulation (Salzman et al. 1990; Celebrini and Newsome 1995; Britten and Wezel 1998; DeAngelis et al. 1998) and not in the absence of visual inputs, as could be in the case of blindness. Groundbreaking studies by Tehovnik and colleagues demonstrated electrical stimulation can evoke sensation of spots or regions of light, termed phosphenes (Tehovinik et al. 2005; Tehovnik and Slocum 2007a, 2007b). However, these methods failed to elicit feature-specific (e.g., orientation-specific) activation. We believe the primary reason for this failure was due to electrical current spread in brain tissue, something which would result in activation of multiple functional domains of differing specificity, resulting in activation that fail to be feature specific. Thus far, because such stimulation studies have not been guided by visualization of cortical response, electrical stimulation has not achieved feature-selective network activation in vivo (cf. Weliky et al. 1995, for combined focal electrical stimulation with electrophysiological recordings in vitro). Two key advances contributed to our success here: selective targeting of domains and achieving focal stimulation. Previous studies had used either surface electrodes which stimulated large areas of cortex, or penetrating microelectrodes but without knowledge of functional domain location. Here, the application of optical imaging methodology to map functional domain locations provided a novel method of targeted electrical stimulation. Optical imaging provided a method to systematically assess the spatial, temporal, and amplitude effects of the stimulation and led to identification of desired parameters. These parameters (biphasic pulses, 200–500 μs/phase, 200 Hz, 150–200 ms train, current amplitude 25–100 μA) achieved activation confined to single functional domains. Another key advance offered by optical imaging centered on the utility of the maps for electrode targeting; this, in turn, led to identification of connected domains of matching orientation. The upshot is that this combined optical and electrical paradigm successfully elicited orientation network specific activation similar to patterns evoked by visual stimulation. Stimulation Paradigm Fulfills Desired Criteria Similar Response Amplitude and Intensity Dependence By comparing the responses evoked by electrical stimulation and visual stimulation, we found that they shared similar response amplitude (Figs 4 and 5). Importantly, the evoked response amplitudes were current intensity dependent, fulfilling an important requirement for an effective visual cortical prosthetic (VCP) (e.g., for mimicking visual contrast response, Albrecht and Hamilton 1982; Zhan et al. 2005). Focal Effect We found that the size of activation elicited by electrical stimulation (current amplitude: 50–100 μA) was around 630–850 μm (see Fig. 6C). This size range is consistent with that of Tehovnik and colleagues. They estimated that 50–100 μA current levels (with 0.5-1 MΩ electrodes) lead to tissue activation sizes of 0.572–0.736 mm in diameter (Tehovnik et al. 2006), a size close to that of one orientation domain in cat area 18 (Swindale et al. 1987; Shmuel and Grinvald 1996; Wang et al. 2011). Thus, both our imaging evidence and modeling estimates support the inference that current can be restricted to single orientation domains. Similar Spatial Pattern In our experiments, the electrodes were inserted to around 400 μm from brain surface, a depth which corresponds to layer 2/3. At this depth, previous studies have shown that there are horizontal connections linking domains of similar orientation preference in cat visual cortex area 18 (Matsubara et al. 1987; Gilbert and Wiesel 1989). The majority of projecting patches were at a distance of around 1 mm away, while long-distance connections extend up to about 3 mm (Matsubara et al. 1987). Consistent with this anatomical connectivity pattern, electrical stimulation activated several patches around the stimulation site (see Fig. 3), most located close to the stimulation site (distances around 1 mm), and some located as far as 2.5 mm away (see Fig. S5). Thus, our results show that electrical stimulation effectively elicits activation in nearby orientation domains of similar orientation preference; these response patterns were stable and repeatable (Fig. 4). Achieves Desired Range of Orientation Encoding In visual cortex, all visual orientations are represented in a systematically shifting fashion across different orientation columns (Hubel and Wiesel 1977). Previous studies have shown, using four different stimulus orientations, that contour orientations can be decoded from evoked optical imaging maps (Lu et al. 2010; Chen et al. 2016; Lu et al. 2018). Here, we have shown that electrical response maps evoked by targeted orientation domain stimulation can be similarly decoded (Fig. 3). The classification results of the response patterns under electrical stimulation are in agreement with the visual decoded predictions (see Fig. S2, Fig. 3 M-O). Thus our results demonstrate that, with appropriate current levels and stimulation locations, it is possible to construct a stable response that is similar in amplitude and pattern to visually evoked cortical responses. Possible Neural Effects of Focal Electrical Stimulation One of the interesting observations here is that stimulating a single orientation domain leads to a response enhancement of orientation-matched domains and a slight reduction of response at orthogonal domains. In effect, using the parameters in this study, increasing focal stimulation of a single orientation domain, leads to increasing differential response between opposing orientation networks. This orientation-selective effect recalls the phenomenon of ‘cross-orientation suppression,” a suppression of orientation-selective response by addition of an orthogonal orientation (Morrone et al. 1982; DeAngelis et al. 1992), one which depends on the relative strength between two orthogonal orientation signals. The mechanism underlying cross-orientation interaction is controversial, as some studies have pointed to contributions from thalamic origins (Freeman et al. 2002; Priebe and Ferster 2006) while others suggest the role of excitatory and inhibitory intracortical circuits (e.g., Kisvárday et al. 1997; Nassi et al. 2015). Previous optical imaging and electrophysiology studies have shown that local horizontal connections mediate both facilitatory and suppressive synaptic interactions largely between orientation-matched columns (Weliky et al. 1995; Toth et al. 1996). How effects on orthogonal domains arise are still under study; however, cortical inhibitory neurons, which have been shown to contact a broad range, including both matched and nonmatched, orientations (e.g., basket cells, Kisvárday and Eysel 1993), could play a potential role. The possible circuits are complex, as there are likely also thalamic and feedback effects of this focal stimulation, possibilities that will require further investigation. Towards a Targeted, Functional Human Brain Machine Interface For VCP, although a substantial number of experiments were conducted on nonhuman primates and several prototype devices have been tested (Brindley and Lewin 1968; Dobelle and Mladejovsky 1974; Dobelle et al. 1974; Dobelle et al. 1976; Bak et al. 1990; Newsome et al. 1990; Salzman et al. 1992; Salzman and Newsome 1994; Schmidt et al. 1996; Dobelle 2000; Cohen and Newsome 2004), none of these VCP have shown that electrical stimulation evokes responses similar to visual stimulation; neither have these studies taken advantage of the functional organization of the cortex. Similar to nonhuman primates and cats, human visual cortex is also comprised of functional domains dedicated to processing these different visual features. For example, primary visual cortex in humans has similar columnar organization such as ocular dominance columns, orientation columns, and color-processing blobs. As such domains in humans are roughly 0.7–1 mm in size (Yacoub et al. 2008; Kuriki et al. 2015), we predict the electrical stimulation paradigm developed here can be confined to single domains to achieve selective visual effects. For people with blindness or severe visual impairment, visual input from retina to cortex is weak and thus the response differences between different functional domains would be weak. Our results suggest that, at least in reasonably intact visual cortex, the introduction of targeted focal microstimulation into cortex would increase the response differences between different functional networks, serving to sharpen selectivity. Notes We thank Y. Liu, Q.N. Wang for their valuable technical assistance during the experiments. We thank R. Friedman for comments on this manuscript. Conflict of Interest: The authors declare no conflict of interest. Author Contributions J.M.H and A.W.R. designed the experiment. J.M.H. and M.Z.Q. collected the data. J.M.H. analyzed the data. A.W.R., J.M.H., X. M. S., and H.T. wrote the manuscript. Funding National Key R&D Program of China (2018YFA0701400 to A.W.R.); National Natural Science Foundation of China (8191101288 and 31627802 to A.W.R.); Key research and development program of Zhejiang province (2020C03004 to A.W.R.). References Albrecht DG , Hamilton DB. 1982 . Striate cortex of monkey and cat: contrast response function . J Neurophysiol. 48 : 217 – 237 . Google Scholar Crossref Search ADS PubMed WorldCat Albus K , Reckmann R. 1980 . Second and third visual areas of the cat: interindividual variability in retinotopic arrangement and cortical location . J Physiol. 299 : 247 – 276 . Google Scholar Crossref Search ADS PubMed WorldCat Bak MJ , Grivin JP, Hambrecht FT, Kufta CV. 1990 . Visual sensations produced by intracortical microstimulation of the human occipital cortex . Med Biol Eng Comput. 28 : 257 – 259 . Google Scholar Crossref Search ADS PubMed WorldCat Bonhoeffer T , Grinvald A. 1996 . Optical imaging based on intrinsic signals: the methodology. In: Toga AW, Mazziotta JC, editors. Brain mapping: the methods . San Diego, CA: Academic Press , pp. 55 – 97 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bosking WH , Beauchamp MS, Yoshor D. 2017 . Electrical stimulation of visual cortex: relevance for the development of visual cortical prosthetics . Annu Rev Vis Sci. 3 : 141 – 166 . Google Scholar Crossref Search ADS PubMed WorldCat Bosking WH , Zhang Y, Schofield B, Fitzpatrick D. 1997 . Orientation selectivity and the arrangement of horizontal connections in tree shrew striate cortex . J Neurosci. 17 : 2112 – 2127 . Google Scholar Crossref Search ADS PubMed WorldCat Brindley GS , Lewin WS. 1968 . The visual sensations produced by electrical stimulation of the medial occipital cortex . J Physiol. 194 : 54 – 55 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Britten KH , van Wezel RJA. 1998 . Electrical microstimulation of cortical area MST biases heading perception in monkeys . Nat Neurosci. 1 : 59 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat Celebrini S , Newsome WT. 1995 . Microstimulation of extrastriate area MSTinfluences performance on a direction discrimination task . J Neurophysiol. 73 : 437 – 448 . Google Scholar Crossref Search ADS PubMed WorldCat Chen G , Lu HD, Roe AW. 2008 . A map of horizontal disparity in primate V2 . Neuron. 58 : 442 – 450 . Google Scholar Crossref Search ADS PubMed WorldCat Chen M , Li P, Zhu S, Han C, Xu H, Fang Y, Hu J, Roe AW, Lu HD. 2016 . An orientation map for motion boundaries in macaque V2 . Cereb Cortex. 26 : 279 – 287 . Google Scholar Crossref Search ADS PubMed WorldCat Cohen MR , Newsome WT. 2004 . What electrical microstimulation has revealed about the neural basis of cognition . Curr Opin Neurobiol. 14 : 1 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat DeAngelis GC , Cumming BG, Newsome WT. 1998 . Cortical area MT and the perception of stereoscopic depth . Nature. 394 : 677 – 680 . Google Scholar Crossref Search ADS PubMed WorldCat DeAngelis GC , Robson JG, Ohzawa I, Freeman RD. 1992 . Organization of suppression in receptive fields of neurons in cat visual cortex . J Neurophysiol. 68 : 144 – 163 . Google Scholar Crossref Search ADS PubMed WorldCat Dobelle WH . 2000 . Artificial vision for the blind by connecting a television camera to the visual cortex . ASAIO J. 46 : 3 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Dobelle WH , Mladejovsky MG. 1974 . Phosphenes produced by electrical stimulation of human occipital cortex, and their application to the development of a prosthesis for the blind . J Physiol. 243 : 553 – 576 . Google Scholar Crossref Search ADS PubMed WorldCat Dobelle WH , Mladejovsky M, Evans JR, Roberts TS. 1976 . “Braille” reading by a blind volunteer by visual cortex stimulation . Nature. 259 : 111 – 112 . Google Scholar Crossref Search ADS PubMed WorldCat Dobelle WH , Mladejovsky MG, Girvin JP. 1974 . Artificial vision for the blind: electrical stimulation of visual cortex offers hope for a functional prosthesis . Science. 183 : 440 – 444 . Google Scholar Crossref Search ADS PubMed WorldCat Fang Y , Chen M, Xu H, Li P, Han C, Hu J, Zhu S, Ma H, Lu HD. 2019 . An orientation map for disparity-defined edges in area V4 . Cereb Cortex. 29 : 666 – 679 . Google Scholar Crossref Search ADS PubMed WorldCat Freeman TCB , Durand S, Kiper DC, Carandini M. 2002 . Suppression without inhibition in visual cortex . Neuron. 35 : 759 – 771 . Google Scholar Crossref Search ADS PubMed WorldCat Ghovanloo M , Najafi K. 2004 . A modular 32-site wireless neural stimulation microsystem . IEEE J Solid-State Circuits. 39 : 2457 – 2466 . Google Scholar Crossref Search ADS WorldCat Ghovanloo M , Najafi K. 2007 . A wireless implantable multichannel microstimulating system-on-a-chip with modular architecture . IEEE Trans Neural Syst Rehabil Eng. 15 : 449 – 457 . Google Scholar Crossref Search ADS PubMed WorldCat Gilbert CD , Wiesel TN. 1989 . Columnar specificity of intrinsic horizontal and corticocortical connections in cat visual cortex . J Neurosci. 9 : 2432 – 2442 . Google Scholar Crossref Search ADS PubMed WorldCat Grinvald A , Lieke E, Frostig RD, Gilbert CD, Wiesel TN. 1986 . Functional architecture of cortex revealed by optical imaging of intrinsic signals . Nature. 324 : 361 – 364 . Google Scholar Crossref Search ADS PubMed WorldCat Grinvald A , Slovin H, Vanzetta I. 2000 . Non-invasive visualization of cortical columns by fMRI . Nature Neurosci. 3 : 105 – 107 . Google Scholar Crossref Search ADS WorldCat Hammond P , Andrews DP. 1978 . Orientation tuning of cells in areas 17 and 18 of the cat's visual cortex . Exp Brain Res. 31 : 341 – 351 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Histed MH , Bonin V, Reid RC. 2009 . Direct activation of sparse, distributed populations of cortical neurons by electrical microstimulation . Neuron. 63 : 508 – 522 . Google Scholar Crossref Search ADS PubMed WorldCat Hochberg LR , Bacher D, Jarosiewicz B, Masse NY, Simeral JD, Vogel J, Haddadin S, Liu J, Cash SS, van der Smagt P et al. 2012 . Reach and grasp by people with tetraplegia using a neurally controlled robotic arm . Nature. 485 : 372 – 375 . Google Scholar Crossref Search ADS PubMed WorldCat Hu J , Ma H, Zhu S, Li P, Xu H, Fang Y, Chen M, Han C, Fang C, Cai X et al. 2018 . Visual motion processing in macaque V2 . Cell Rep. 25 : 157 – 167 . Google Scholar Crossref Search ADS PubMed WorldCat Hubel DH , Wiesel TN. 1977 . Functional architecture of macaque monkey visual cortex . Proc R Soc Lond B Biol Sci. 198 : 1 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat Kisvárday ZF , Eysel UT. 1993 . Functional and structural topography of horizontal inhibitory connections in cat visual cortex . Eur J Neurosci. 5 : 1558 – 1572 . Google Scholar Crossref Search ADS PubMed WorldCat Kisvárday ZF , Toth E, Rausch M, Eysel UT. 1997 . Orientation-specific relationship between populations of excitatory and inhibitory lateral connections in the visual cortex of the cat . Cereb Cortex. 7 : 605 – 618 . Google Scholar Crossref Search ADS PubMed WorldCat Kuriki I , Sun P, Ueno K, Tanaka K, Cheng K. 2015 . Hue selectivity in human visual cortex revealed by functional magnetic resonance imaging . Cereb Cortex. 25 : 4869 – 4884 . Google Scholar Crossref Search ADS PubMed WorldCat Levitt JB , Yoshioka T, Lund JS. 1994 . Intrinsic cortical connections in macaque visual area V2: evidence for interaction between different functional streams . J Comp Neurol. 342 : 551 – 570 . Google Scholar Crossref Search ADS PubMed WorldCat Lewis PM , Rosenfeld JV. 2016 . Electrical stimulation of the brain and the development of cortical visual prostheses: an historical perspective . Brain Res. 1630 : 208 – 224 . Google Scholar Crossref Search ADS PubMed WorldCat Li P , Zhu S, Chen M, Han C, Xu H, Hu J, Fang Y, Lu HD. 2013 . A motion direction preference map in monkey V4 . Neuron. 78 : 376 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat Li Z , Meng J, Li H, Jin A, Tang Q, Zhu J, Yu H. 2017 . The feature-specific propagation of orientation and direction adaptation from areas 17 to 21a in cats . Sci Rep. 7 : 390 . Google Scholar Crossref Search ADS PubMed WorldCat Lu HD , Chen G, Cai J, Roe AW. 2017 . Intrinsic signal optical imaging of visual brain activity: tracking of fast cortical dynamics . Neuroimage. 148 : 160 – 168 . Google Scholar Crossref Search ADS PubMed WorldCat Lu HD , Chen G, Tanigaw H, Roe AW. 2010 . A motion direction map in macaque V2 . Neuron. 68 : 1002 – 1013 . Google Scholar Crossref Search ADS PubMed WorldCat Lu HD , Roe AW. 2008 . Functional organization of color domains in V1 and V2 of macaque monkey revealed by optical imaging . Cereb Cortex. 18 : 516 – 533 . Google Scholar Crossref Search ADS PubMed WorldCat Lu YL , Yin JP, Chen ZY, Gong HL, Liu Y, Qian LL, Li XH, Liu R, Andolina IM, Wang W. 2018 . Revealing detail along the visual hierarchy: neural clustering preserves acuity from V1 to V4 . Neuron. 98 : 417 – 428 . Google Scholar Crossref Search ADS PubMed WorldCat Matsubara JA , Cynader MS, Swindale NV. 1987 . Anatomical properties and physiological correlates of the intrinsic connections in cat area 18 . J Neurosci. 7 : 1428 – 1446 . Google Scholar Crossref Search ADS PubMed WorldCat Moeller S , Crapse T, Tsao DY. 2017 . The effect of face patch microstimulation on perception of faces and objects . Nat Neurosci. 20 : 743 – 754 . Google Scholar Crossref Search ADS PubMed WorldCat Morrone MC , Burr DC, Maffei L. 1982 . Functional implications of crossorientation inhibition of cortical visual cells. I. Neurophysiological evidence . Proc R Soc Lond B Biol Sci. 216 : 335 – 354 . Google Scholar Crossref Search ADS PubMed WorldCat Murasugi CM , Salzman CD, Newsome WT. 1993 . Microstimulation in visual area MT: effects of varying pulse amplitude and frequency . J Neurosci. 13 : 1719 – 1729 . Google Scholar Crossref Search ADS PubMed WorldCat Murphey DK , Maunsell JH. 2007 . Behavioral detection of electrical microstimulation in different cortical visual areas . Curr Biol. 17 : 862 – 867 . Google Scholar Crossref Search ADS PubMed WorldCat Nassi JJ , Avery MC, Cetin AH, Roe AW, Reynolds JH. 2015 . Optogenetic activation of normalization in alert macaque visual cortex . Neuron. 86 : 1504 – 1517 . Google Scholar Crossref Search ADS PubMed WorldCat Newsome WT , Britten KH, Salzman CD, Movshon JA. 1990 . Neuronal mechanisms of motion perception . Cold Spring Harb Symp Quant Biol. 55 : 697 – 705 . Google Scholar Crossref Search ADS PubMed WorldCat Penfield W , Perot P. 1963 . The brain’s record of auditory and visual experience. A final summary and discussion . Brain. 86 : 595 – 696 . Google Scholar Crossref Search ADS PubMed WorldCat Priebe NJ , Ferster D. 2006 . Mechanisms underlying cross-orientation suppression in cat visual cortex . Nat Neurosci. 9 : 552 – 561 . Google Scholar Crossref Search ADS PubMed WorldCat Ranck JB . 1975 . Which elements are excited in electrical stimulation of mammalian central nervous system: a review . Brain Res. 98 : 417 – 440 . Google Scholar Crossref Search ADS PubMed WorldCat Roe AW . 2007 . Long-term optical imaging of intrinsic signals in anesthetized and awake monkeys . Appl Optics. 46 : 1872 – 1880 . Google Scholar Crossref Search ADS WorldCat Roe AW , Chelazzi L, Connor CE, Conway BR, Fujita I, Gallant J, Lu H, Vanduffel W. 2012 . Towards a unified theory of visual area V4 . Neuron. 74 : 12 – 29 . Google Scholar Crossref Search ADS PubMed WorldCat Romo R , Hernandez A, Zainos A, Salinas E. 1998 . Somatosensory discrimination based on cortical microstimulation . Nature. 392 : 387 – 390 . Google Scholar Crossref Search ADS PubMed WorldCat Salzman CD , Britten KH, Newsome WT. 1990 . Cortical microstimulation influences perceptual judgements of motion direction . Nature. 346 : 174 – 177 . Google Scholar Crossref Search ADS PubMed WorldCat Salzman CD , Murasugi CM, Britten KH, Newsome WT. 1992 . Microstimulation in visual area MT: effects on direction discrimination performance . J Neurosci. 12 : 2331 – 2355 . Google Scholar Crossref Search ADS PubMed WorldCat Salzman CD , Newsome WT. 1994 . Neural mechanisms for forming a perceptual decision . Science. 264 : 231 – 237 . Google Scholar Crossref Search ADS PubMed WorldCat Schmidt EM , Bak MJ, Hambrecht FT, Kufta CV, O’Rourke DK, Vallabhanath P. 1996 . Feasibility of a visual prosthesis for the blind based on intracortical microstimulation of the visual cortex . Brain. 119 : 507 – 522 . Google Scholar Crossref Search ADS PubMed WorldCat Shmuel A , Grinvald A. 1996 . Functional organization for direction of motion and its relationship to orientation maps in cat area 18 . J Neurosci. 16 : 6945 – 6964 . Google Scholar Crossref Search ADS PubMed WorldCat Sincich LC , Blasdel GG. 2001 . Oriented axon projections in primary visual cortex of the monkey . J Neurosci. 21 : 4416 – 4426 . Google Scholar Crossref Search ADS PubMed WorldCat Stepniewska I , Friedman R, Gharbawie OA, Cerkevich CM, Roe AW, Kaas JH. 2011 . Optical imaging in galagos reveals parietal-frontal circuits underlying motor behavior . Proc Natl Acad Sci. 37 : 15033 – 15034 . OpenURL Placeholder Text WorldCat Stettler DD , Das A, Bennett J, Gilbert CD. 2002 . Lateral connectivity and contextual interactions in macaque primary visual cortex . Neuron. 36 : 739 – 750 . Google Scholar Crossref Search ADS PubMed WorldCat Swindale NV , Matsubara JA, Cynader MS. 1987 . Surface organization of orientation and direction selectivity in cat area 18 . J Neurosci. 7 : 1414 – 1427 . Google Scholar Crossref Search ADS PubMed WorldCat Tanigawa H , Lu HD, Roe AW. 2010 . Functional organization for color and orientation in macaque V4 . Nat Neurosci. 13 : 1542 – 1135 . Google Scholar Crossref Search ADS PubMed WorldCat Tehovnik EJ . 1996 . Electrical stimulation of neural tissue to evoke behavioral responses . J Neurosci Methods. 65 : 1 – 17 . Google Scholar Crossref Search ADS PubMed WorldCat Tehovnik EJ , Slocum WM. 2007a . Phosphene induction by microstimulation of macaque V1 . Brain Res Rev. 53 : 337 – 343 . Google Scholar Crossref Search ADS WorldCat Tehovnik EJ , Slocum WM. 2007b . What delay fields tell us about striate cortex . J Neurophysiol. 98 : 559 – 576 . Google Scholar Crossref Search ADS WorldCat Tehovnik EJ , Slocum WM, Schiller PH. 2004 . Microstimulation ofV1 delays the execution of visually guided saccades . Eur J Neurosci. 20 : 264 – 272 . Google Scholar Crossref Search ADS PubMed WorldCat Tehovnik EJ , Slocum WM, Schiller PH. 2005 . Phosphene induction and the generation of saccadic eyemovements by striate cortex . J Neurophysiol. 93 : 1 – 19 . Google Scholar Crossref Search ADS PubMed WorldCat Tehovnik EJ , Tolias AS, Sultan F, Slocum WM, Logothetis NK. 2006 . Direct and indirect activation of cortical neurons by electrical microstimulation . J Neurophysiol. 96 : 512 – 521 . Google Scholar Crossref Search ADS PubMed WorldCat Tolias AS , Sultan F, Augath M, Oeltermann A, Tehovnik EJ, Schiller PH, Logothetis NK. 2005 . Mapping cortical activity elicited with electrical microstimulation using fMRI in the macaque . Neuron. 48 : 901 – 911 . Google Scholar Crossref Search ADS PubMed WorldCat Toth LJ , Rao SC, Kim DS, Somers D, Sur M. 1996 . Subthreshold facilitation and suppression in primary visual cortex revealed by intrinsic signal imaging . Proc Natl Acad Sci. 93 : 9869 – 9874 . Google Scholar Crossref Search ADS WorldCat Tusa RJ , Rosenquist AC, Palmer LA. 1979 . Retinotopic organization of Area-18 and Area-19 in the cat . J Comp Neurol. 185 : 657 – 678 . Google Scholar Crossref Search ADS PubMed WorldCat Wang G , Nagai M, Okamura J. 2011 . Orientation dependency of intrinsic optical signal dynamics in cat area 18 . Neuroimage. 57 : 1140 – 1153 . Google Scholar Crossref Search ADS PubMed WorldCat Weliky M , Kandler K, Fitzpatrick D, Katz LC. 1995 . Patterns of excitation and inhibition evoked by horizontal connections in visual cortex share a common relationship to orientation columns . Neuron. 15 : 541 – 552 . Google Scholar Crossref Search ADS PubMed WorldCat Wichmann T , Delong MR. 2006 . Deep brain stimulation for neurologic and neuropsychiatric disorders . Neuron. 52 : 197 – 204 . Google Scholar Crossref Search ADS PubMed WorldCat Yacoub D , Harel N, Ugurbil K. 2008 . High-field fMRI unveils orientation columns in humans . Proc Natl Acad Sci. 105 : 10607 – 10612 . Google Scholar Crossref Search ADS WorldCat Zhan C , Ledgeway T, Baker CL. 2005 . Contrast response in visual cortex: quantitative assessment with intrinsic optical signal imaging and neural firing . Neuroimage. 26 : 330 – 346 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permission@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Focal Electrical Stimulation of Cortical Functional Networks
JF - Cerebral Cortex
DO - 10.1093/cercor/bhaa136
DA - 2019-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/focal-electrical-stimulation-of-cortical-functional-networks-QwDueCNmvy
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -