- Email: [email protected]
A comparison of monkey and human motion processing mechanisms
Vision Research 50 (2010) 2137–2141
Contents lists available at ScienceDirect
Vision Research journal homepage: www.elsevier.com/locate/visres
A comparison of monkey and human motion processing mechanisms Catherine Lynn, William Curran ⇑ School of Psychology, Queen’s University Belfast, Belfast, BT7 1NN, UK
a r t i c l e
i n f o
Article history: Received 4 June 2010 Received in revised form 3 August 2010
Keywords: Motion processing Motion adaptation Direction aftereffect Motion-sensitive neurons
a b s t r a c t Single-cell recording studies have provided vision scientists with a detailed understanding of motion processing at the neuronal level in non-human primates. However, despite the development of brain imaging techniques, it is not known to what extent the response characteristics of motion-sensitive neurons in monkey brain mirror those of human motion-sensitive neurons. Using a motion adaptation paradigm, the direction aftereffect, we recently provided evidence of a strong resemblance in the response functions of motion-sensitive neurons in monkey and human to moving dot patterns differing in dot density. Here we describe a series of experiments in which measurements of the direction aftereffect are used to infer the response characteristics of human motion-sensitive neurons when viewing transparent motion and moving patterns that differ in their signal-to-noise ratio (motion coherence). In the case of transparent motion stimuli, our data suggest suppressed activity of motion-sensitive neurons similar to that reported for macaque monkey. In the case of motion coherence, our results are indicative of a linear relationship between signal intensity (coherence) and neural activity; a pattern of activity which also bears a striking similarity to macaque neural activity. These findings strongly suggest that monkey and human motionsensitive neurons exhibit similar response and inhibitory characteristics. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Single-cell recording studies have provided a detailed characterization of the response properties of motion-sensitive neurons in monkey visual cortex (Albright, 1984; Dubner & Zeki, 1971; Krekelberg & Albright, 2005; Lagae, Gulyas, Raiguel, & Orban, 1989; Newsome & Paré, 1988). This approach, which was instrumental in identifying monkey area MT/V5 as an important cortical site for motion processing, typically relies on neuronal spiking as a measure of neural activity. Indeed, spiking discharge of MT/V5 direction-sensitive neurons has been shown to be causally related to behavioural measurements of direction perception (Ditterich, Mazurek, & Shadlen, 2003; Salzman, Murasugi, Britten, & Newsome, 1992). Functional magnetic resonance imaging (fMRI) has been deployed effectively for the study of motion processing in the human brain (Bartels, Logothetis, & Moutoussis, 2008; Morrone et al., 2000; Tootell et al., 1995; Watson et al., 1993), and has been used to infer the response properties of human motion-sensitive neurons. However, how neuronal spiking influences fMRI signal is, as yet, unclear. While some studies have found fMRI signals to be directly proportional to average neuronal spiking rates (Boynton, Demb, Glover, & Heeger, 1999; Rees, Friston, &
⇑ Corresponding author. E-mail address: [email protected] (W. Curran). 0042-6989/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.visres.2010.08.007
Koch, 2000), recent findings suggest that fMRI signals are most likely driven by synaptic activity rather than neuronal spiking activity (Viswanathan & Freeman, 2007). In a review of what can be inferred from fMRI signals, Logothetis (2008) concludes that it is unrealistic at present to draw a direct analogy between neuronal spiking data from animal experiments and fMRI signals derived from human studies. The spike discharge of direction-sensitive neurons in monkey is determined by a number of factors; including how well the stimulus motion direction is matched to the neurons’ preferred direction (Albright, 1984; Snowden, Treue, & Andersen, 1992), and the ‘strength’ of the motion signal. Motion signal strength can be altered by manipulating stimulus contrast (Sclar, Maunsell, & Lennie, 1990), motion coherence (Britten, Shadlen, Newsome, & Movshon, 1993) or dot density (Snowden, Treue, Erickson, & Andersen, 1991; Snowden et al., 1992). When its preferred direction of motion is presented within a neuron’s receptive field for a prolonged period of time a process known as adaptation occurs in which the neuron’s responsiveness is reduced and its direction tuning is modified. Kohn and Movshon (2004) have shown that adaptation of MT direction-sensitive neurons to near-preferred directions causes direction tuning to shift towards the adapted direction. This direction tuning shift is consistent with a perceptual aftereffect known as the direction aftereffect (see below). Stimuli that evoke a strong response in targeted neurons generally act as more effective adaptors (Kohn, 2007; van Wezel & Britten, 2002). More effective motion
2138
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
adaptors, in turn, induce greater changes in neurons’ direction tuning functions, which are thought to impact on the magnitude of motion-adaptation phenomena (Kohn, 2007; Kohn & Movshon, 2004). Given what we know about the response characteristics of monkey direction-sensitive neurons, and how their response functions change following adaptation, it is possible to measure (indirectly) whether human direction-sensitive neurons have similar or different response characteristics to monkey. This can be achieved through the use of motion-adaptation phenomena, such as the direction aftereffect (DAE). The DAE describes the misperception of a motion direction following prolonged viewing of (adaptation to) a different direction of motion (Curran, Clifford, & Benton, 2006a; Levinson & Sekuler, 1976; Patterson & Becker, 1996; Wiese & Wenderoth, 2007). While it has received considerably less attention than the well known motion aftereffect (Mather, Verstraten, & Anstis, 1998), the DAE has contributed to our understanding of speed and direction coding. For example, it has been used to support previous reports of the motion system having at least two speed channels (Curran, Clifford, & Benton, 2006b). Furthermore the DAE and its simultaneous counterpart, direction repulsion (Marshak & Sekuler, 1979; Mather & Moulden, 1980), have been cited as strong evidence for the distribution-shift model of direction coding (Mather & Moulden, 1980); a model which proposes that direction is coded by the population response distribution of motion-sensitive neurons. If motion-sensitive neurons have similar response characteristics in human and monkey, and given the reported relationship between neural spiking and aftereffect magnitude, motion stimuli that are known to have differential effects on the spiking discharge of monkey direction-sensitive neurons should induce DAEs that differ in magnitude. Varying dot density is a prime example of a motion stimulus manipulation which differentially affects neural spiking. Snowden et al. (1991, 1992) reported that macaque motion-sensitive neurons rapidly increase their response to their preferred direction as the number of moving dots within their receptive fields increases, and that this rapid increase in spike discharge plateaus at relatively low dot densities (at approximately 8 dots/deg2). When additional dots moving in the neurons’ anti-preferred direction were added, Snowden et al. (1991) recorded density-tuning functions indicative of a division-like inhibitory interaction between neurons tuned to opposite directions. In a recent study Curran and Lynn (2009) used the DAE to investigate whether human motion-sensitive neurons respond to varying dot density in a similar manner to macaque. It was reasoned that, if the differential responses to dot density found in macaque mirror the response characteristics of human motion-sensitive neurons, low dot density stimuli would be less effective adaptors than high dot density stimuli. Given the reported relationship between neural spiking and aftereffect magnitude (Kohn, 2007; van Wezel & Britten, 2002), the former stimuli should induce a weaker DAE than the latter stimuli. The results were strikingly similar to Snowden et al.’s (1991) single-cell recording data; DAE magnitude increased rapidly with increasing adaptor dot density up to 10 dots/deg2, then levelled off and remained constant for higher dot densities. When the experiment was repeated, but this time with additional dots moving in the opposite direction, the DAE density-tuning functions were consistent with division-like inhibitory interactions between neurons tuned to opposite directions. Furthermore, when equal numbers of dots moved in opposite directions a measurable DAE was still evident; this concurs with Snowden et al.’s (1991) finding that direction-sensitive neurons continue to respond (albeit at a reduced rate) when equal numbers of dots move in the neurons’ preferred and anti-preferred directions. These results strongly suggest that, in the limited case of dot density, response charac-
teristics of macaque motion-sensitive neurons mirror those of human motion-sensitive neurons. The current series of experiments investigates further the extent to which the response characteristics of macaque motionsensitive neurons are an accurate model of human motion processing. Experiment 1 uses DAE magnitude measurements to infer whether the responsiveness of direction-sensitive neurons is suppressed when presented with two transparent motions moving in two distinct directions. Snowden et al. (1991) have previously addressed this question with respect to macaque motion-sensitive neurons. In their experiment they identified the preferred direction for a number of MT neurons; then recorded the neurons’ responses when two superimposed, transparent motions were presented within a neuron’s receptive field. One of the transparent motions moved in the neuron’s preferred direction, and the second motion direction differed from the first by up to 180° (in steps of 45°). Snowden et al. observed suppression, but not complete extinction, of neural spike discharge under the transparent motions conditions. Interestingly, two of the non-preferred directions (45° and 90°) produced an excitatory response when presented in isolation, despite the fact that they produced an inhibitory effect when presented with the preferred direction. Treue, Hol, and Rauber (2000) found similar suppressive effects. These authors recorded the spike discharge of MT neurons as a function of the direction difference between two transparent motions, and as a function of the mean direction of the transparent motions relative to the neuron’s preferred direction. As in the Snowden et al. paper, Treue et al’s data reveal a clear suppressive effect on spike discharge when one of the two transparent motions matched the neuron’s preferred direction. Both Snowden et al., and Treue et al., report that the suppressive effect increases with increasing direction difference between the transparent motions; and that this increasing suppression either levels out (Treue et al., 2000) or slows dramatically (Snowden et al., 1991) for direction differences greater than 90°. If macaque motion-sensitive neuron activity is an accurate model for human motion-sensitive neurons, the macaque data generated by the above studies make clear predictions for DAE measurements in Experiment 1. Two transparent motions that differ in their directions should suppress the responses of neurons preferentially tuned to either of the two directions. Given the relationship between neural spiking and aftereffect magnitude, adapting to such a stimulus should produce a weaker DAE than adapting to one of the directions in isolation. While DAE magnitude will be reduced under transparent motion conditions, the macaque data dictate that it will not be extinguished; but, rather, should level out for transparent adaptors whose component direction difference exceeds 90°. Experiment 2 uses DAE magnitude measurements to infer the relationship between motion coherence strength and neural spiking activity. Motion coherence strength in a random dot display is determined by the proportion of dots that move in a single, coherent direction. In the case of 0% motion coherence, all dots are randomly re-plotted for each frame refresh. In a stimulus with 50% motion coherence, half the dots are randomly replotted for each frame and half are re-plotted with a specific spatiotemporal relation to dots displayed in the previous frame. Britten et al. (1993) recorded the spike discharge of macaque MT neurons as a function of the motion coherence of stimuli presented within the neurons’ receptive fields. They found that, for most of the neurons, spike discharge increased linearly as a function of motion coherence level. If macaque and human motion-sensitive neurons respond similarly to changes in motion coherence, Britten et al’s data predict that DAE magnitude should increase linearly with increasing motion coherence of the adapting stimulus.
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
2. Experiment 1. Responses of human motion-sensitive neurons to transparent motion mirror those of macaque 2.1. Observers Four observers – one of the authors and three naïve participants – took part in the experiment.
2139
vertically upwards. DAE magnitude was taken as the difference between the two directions (perceived and actual). A range of direction differences was used for the transparent adaptor stimuli (30°, 60°, 90° and 120°). Observers generated four psychometric functions each for the baseline and transparency conditions, with each psychometric function being derived from 64 trials. The interval between testing with different adaptors was at least 15 min, thus ensuring recovery from adaptation.
2.2. Apparatus and stimuli 3. Results Stimuli were displayed on a Mitsubishi 2070SB monitor, and comprised random dot kinematograms (RDK) presented within a circular aperture (7 deg2), with each RDK containing equal numbers of black and white dots (dot diameter = 1.8 arcmin) against a mean luminance background (59 cd per m2). The monitor was driven by a Cambridge Research Systems Visage graphics board at a frame rate of 80 Hz. 2.3. Procedure During the initial motion adaptation phase (30 s duration) observers were presented with a random dot adapting stimulus, in which either all dots moved 45° to the right of vertical upwards (baseline condition) or half the dots moved in the baseline direction and remaining dots moved in a direction clockwise from 45° (transparency condition). Adaptor speed was fixed at 2.5° s 1, and adaptor dot density was set to either 28 dots/deg2 (baseline condition) or 56 dots/deg2 (transparency condition). The adaptor direction(s) was the same for all subsequent top-up phases. To help maintain fixation, a central spot was present throughout each experimental run. In the test phase following adaptation, observers judged whether the test stimulus (speed 2.5° s 1, dot density 28 dots/deg2, duration 200 ms) was moving left or right of vertical up. A vertical line extended from the top and bottom of the test stimulus aperture, thus providing observers with a reference to vertical. In order to maintain adaptation, test phases alternated with adaptation top-up phases of 5 s duration. Test stimulus motion direction was chosen by an adaptive method of constant stimuli (adaptive probit estimation), a method that dynamically updates the set of stimulus motion directions being presented depending on the observer’s previous responses (Treutwein, 1995). The stimulus values are selected to optimize the estimation of the point of subjective equality (PSE), in our case the direction the test stimulus was moving when it was perceived as moving
Fig. 1a shows the same pattern of results for all observers. DAE magnitude is greatest for the baseline condition, when all adaptor dots move in the same direction (45°). Under the transparent adaptor conditions, when half the dots move in the baseline direction and half move in a direction clockwise of the baseline direction, there is clear evidence of DAE magnitude decreasing with increasing direction difference between the transparent motion components. This decrease in DAE magnitude continues for direction differences up to 60°, and levels out for larger adaptor direction differences. t-Test analysis reveals a significant reduction in DAE magnitude between the 0° and 30° conditions (df = 15, t = 3.914, p < 0.001), and between the 30° and 60° conditions (df = 15, t = 4.94, p < 0.001). The data suggest that activity of neural populations sensitive to the baseline direction is suppressed when the adaptor stimulus contains an additional direction of motion. This is consistent with macaque data (Snowden et al., 1991), and suggests that human and macaque motion-sensitive neurons respond similarly to transparently moving stimuli. It is interesting to note that, whereas Snowden et al. report a reduction of approximately 70% in neural firing when the preferred and additional motions differ by 90°, DAE magnitude is reduced by approximately 25% under the same condition. This suggests that the relationship between neural activity level and DAE magnitude may not be a straightforward linear one. Snowden et al. (1991) reported that motion directions differing by up to 90° from a neuron’s preferred direction suppressed the neuron’s spike discharge when presented in combination with the preferred motion direction. The suppressive effect developed quickly up to a direction difference of 90° and increased very slowly for larger direction differences. Treue et al.’s (2000) data, on the other hand, reveal a levelling off of the suppressive effect for direction differences greater than 90°. Snowden et al. also found that a subset of suppressive motion directions – up to 90° –
Fig. 1. (a) Direction aftereffect magnitude following adaptation to either a single motion moving in the baseline direction (45° from vertical) or a transparent motion stimulus in which half the dots move in the baseline direction and half move in a direction clockwise from the baseline direction. Light grey symbols plot individual DAE functions for each of the four observers, and the filled circles plot the mean DAE function. Each symbol is the average of four PSEs. Note that the strongest DAE is generated following adaptation to the baseline direction. When dots moving in a different direction are added to the adapting stimulus, DAE magnitude initially decreases with increasing direction difference before levelling out and remaining constant for direction differences of 60° or more. Interestingly, following adaptation to each of the individual directions (b), a DAE is induced by directions 30° and 60° from the baseline direction; yet these same directions have a suppressive effect on the DAE when combined with the baseline direction. Error bars represent ±1 standard error.
2140
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
produced an excitatory response when presented alone. To ascertain whether this finding also applies to human motion processing, we repeated our experiment; but this time had observers adapt to single motion directions (dot density = 28 dots/deg2) ranging from 30° to 120° clockwise from the baseline direction (45°). The results (Fig. 1b) show that adaptors whose directions differed by up to 60° from the baseline direction of Experiment 1 induce a measurable DAE in test stimuli moving vertically up; this is despite the fact that these same directions suppress DAE magnitude when presented with the baseline direction. Statistical analysis shows the effect to be significant (30° from baseline: df = 15, t = 23.86, p < .001; 60° from baseline: df = 15, t = 7.402, p < .001). 4. Experiment 2. Responses of human motion-sensitive neurons to varying motion coherence mirror those of macaque This experiment investigated the effect that varying motion coherence has on DAE magnitude. If human and macaque motion-sensitive neurons respond similarly to motion coherence manipulation, previous single-cell recording data (Britten et al., 1993) would predict a linear relationship between an adapting stimulus’s motion coherence level and DAE magnitude. 4.1. Observers Five observers – one of the authors and four naïve – participated in the experiment. 4.2. Stimuli As in Experiment 1, a random dot display was used for adaptor and test stimuli. The adaptor stimulus’s motion coherence was varied between 0% (all dots randomly re-plotted on each frame) and 100% (all dots moved in the direction 45° right of vertical upwards). For any given motion coherence level, all dots had the same probability of being labelled a signal dot in each frame. The probability that any dot would continue being a signal dot through N frames was cN 1, where c signifies the proportion of coherently moving dots. The test stimulus always contained 100% motion coherence, in which all dots moved in a uniform direction. Adaptor and test stimuli had identical dot densities (64 dots/deg2). 4.3. Procedure Apart from the differences between adaptor stimuli used in Experiments 1 and 2, the procedure was the same as in Experiment 1. Six motion coherence levels were used for the adaptor stimuli – 0%, 20%, 40%, 60%, 80% and 100%. Observers generated four psychometric functions per coherence condition, and took a break of at least 15 min between each adaptor condition to allow recovery from adaptation. 4.4. Results The results from this experiment reveal a strongly linear relationship between adaptor coherence level and DAE magnitude (Fig. 2). As in Experiment 1, these data are strongly consistent with what one would expect if human and macaque motion-sensitive neurons share similar response characteristics. 5. Discussion Much of what we know about the response characteristics of motion-sensitive neurons is derived from single-cell recordings in non-human primates. If one is to develop accurate models of human
Fig. 2. Results from Experiment 2, in which DAE magnitude was measured as a function of the adaptor stimulus’s motion coherence. The data reveal a clear linear relationship between motion coherence level and DAE magnitude.
motion processing, algorithms implemented by such models should be constrained by the known physiological properties of the visual system. To this end, monkey single-cell recording data have played an important role in informing the development of motion models. Implicit in this approach is the assumption that the response characteristics of monkey motion-sensitive neurons mirror faithfully the response characteristics of motion-sensitive neurons in the human brain. The development of brain imaging technology, particularly fMRI, has permitted researchers to gauge the extent to which the response characteristics of human motion-processing neurons mirror those of monkey. Indeed, a number of fMRI studies have revealed striking similarities between human fMRI BOLD signal and monkey single-cell recording data (Huk, Rees, & Heeger, 2001; Krekelberg, Boynton, & van Wezel, 2006; Rees et al., 2000). However, the reported coupling between local field potentials and changes in tissue oxygenation levels (Viswanathan & Freeman, 2007) suggests that it is currently unrealistic to draw a direct analogy between the fMRI BOLD signal from human recordings and neuronal spiking data obtained from monkey experiments (Logothetis, 2008). Thus it is not yet clear to what extent the spiking activity of motion-sensitive neurons in monkey accurately reflects the spiking activity of human motion-sensitive neurons. We used a well known motion-adaptation phenomenon, the direction aftereffect, to infer changes in the spiking activity of motion-sensitive neurons as a function of manipulating transparent motion directions and motion coherence. This indirect approach to inferring neural responsiveness was motivated by the observations that: (1) stimuli that evoke a more robust response in targeted neurons generally act as more effective adaptors (van Wezel & Britten, 2002), and (2) more effective motion adaptors induce greater changes in neurons’ direction tuning functions, which is thought to impact on DAE magnitude (Kohn, 2007; Kohn & Movshon, 2004). By identifying motion stimulus manipulations that are known to result in changes in spiking rate of monkey motion-sensitive neurons, and using these stimuli as adaptors, one can make clear predictions about the impact of these adapting stimuli on DAE magnitude. If DAE magnitude changes are in line with the predictions based on single-cell recording data, this would be compelling evidence that the response characteristics of monkey motion-sensitive neurons mirror those of human neurons. The experiments described here build on recent research suggesting that monkey and human motion-sensitive neurons show the same differential responses to random dot motion stimuli differing in dot density (Curran & Lynn, 2009). The current experiments have extended this comparison between monkey and
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
human to two other types of motion stimuli – transparent and coherent motion. The results from Experiment 1 provide compelling evidence that, just as in macaque (Snowden et al., 1991; Treue et al., 2000), the responsiveness of direction-sensitive neurons to their preferred direction is suppressed when a second motion direction is superimposed on the neurons’ preferred direction. Furthermore, two of the directions that suppressed DAE magnitude were found to induce a measurable DAE when used as a single direction adaptor. This is what would be predicted given Snowden et al’s finding that additional directions suppress the responsiveness of neurons when superimposed on the neurons’ preferred direction, yet have an excitatory effect when presented alone. In the case of motion coherence, the data from Experiment 2 are indicative of a linear increase in neural responsiveness as motion coherence rises. Again, this is in line with monkey single-cell recording data (Britten et al., 1993) in which the majority of MT neurons display spiking activity that increases in a linear fashion with increasing motion coherence. By using the technique of psychophysical adaptation to make inferences about neural processing in human visual cortex, we demonstrate that human motion perception is qualitatively consistent with the operation of the principal animal model of motion processing – macaque area MT. Thus these experimental results support the common held belief that the properties of monkey motion-sensitive neurons can be used as a guide to explain many aspects of human motion processing. References Albright, T. D. (1984). Direction and orientation selectivity of neurons in visual area MT of the macaque. Journal of Neurophysiology, 52, 1106–1130. Bartels, A., Logothetis, N. K., & Moutoussis, K. (2008). FMRI and its interpretations: An illustration on directional selectivity in area V5/MT. Trends in Neuroscience, 31, 444–453. Boynton, G. M., Demb, J. B., Glover, G. H., & Heeger, D. J. (1999). Neuronal basis of contrast discrimination. Vision Research, 39, 257–269. Britten, K. H., Shadlen, M. N., Newsome, W. T., & Movshon, J. A. (1993). Responses of neurons in macaque MT to stochastic motion signals. Visual Neuroscience, 10, 1157–1169. Curran, W., Clifford, C. W., & Benton, C. P. (2006a). The direction aftereffect is driven by adaptation of local motion detectors. Vision Research, 46, 4270–4278. Curran, W., Clifford, C. W. G., & Benton, C. P. (2006b). New binary direction aftereffect does not add up. Journal of Vision, 6, 1451–1458. Curran, W., & Lynn, C. (2009). Monkey and humans exhibit similar motionprocessing mechanisms. Biological Letters, 5, 743–745. Ditterich, J., Mazurek, M. E., & Shadlen, M. N. (2003). Microstimulation of visual cortex affects the speed of perceptual decisions. Nature Neuroscience, 6, 891–898. Dubner, R., & Zeki, S. M. (1971). Response properties and receptive fields of cells in an anatomically defined region of the superior temporal sulcus in the monkey. Brain Research, 35, 528–532. Huk, A. C., Rees, D., & Heeger, D. J. (2001). Neuronal basis of the motion aftereffect reconsidered. Neuron, 32, 161–172.
2141
Kohn, A. (2007). Visual adaptation: Physiology, mechanisms, and functional benefits. Journal of Neurophysiology, 97, 3155–3164. Kohn, A., & Movshon, J. A. (2004). Adaptation changes the direction tuning of macaque MT neurons. Nature Neuroscience, 7, 764–772. Krekelberg, B., & Albright, T. D. (2005). Motion mechanisms in macaque MT. Journal of Neurophysiology, 93, 2908–2921. Krekelberg, B., Boynton, G. M., & van Wezel, R. J. A. (2006). Adaptation: From single cells to BOLD signals. Trends in Neuroscience, 29, 250–256. Lagae, L., Gulyas, B., Raiguel, S., & Orban, G. A. (1989). Laminar analysis of motion information processing in macaque V5. Brain Research, 496, 361–367. Levinson, E., & Sekuler, R. (1976). Adaptation alters perceived direction of motion. Vision Research, 16, 779–781. Logothetis, N. K. (2008). What we can do and what we cannot do with fMRI. Nature, 453, 869–878. Marshak, W., & Sekuler, R. (1979). Mutual repulsion between moving visual targets. Science, 205, 1399–1401. Mather, G., & Moulden, B. (1980). A simultaneous shift in apparent direction: Further evidence for a ‘‘distribution-shift” model of direction coding. Quarterly Journal of Experimental Psychology, 32, 325–333. Mather, G., Verstraten, F., & Anstis, S. (1998). The motion aftereffect. A modern perspective. Cambridge, Massuchusetts: MIT Press. Morrone, M. C., Tosetti, M., Montanaro, D., Fiorentini, A., Cioni, G., & Burr, D. C. (2000). A cortical area that responds specifically to optic flow, revealed by fMRI. Nature Neuroscience, 3, 1322–1328. Newsome, W. T., & Paré, E. B. (1988). A selective impairment of motion perception following lesions of the middle temporal visual area (MT). Journal of Neuroscience, 8, 2201–2211. Patterson, R., & Becker, S. (1996). Direction-selective adaptation and simultaneous contrast induced by stereoscopic (cyclopean) motion. Vision Research, 36, 1773–1781. Rees, G., Friston, K., & Koch, C. (2000). A direct quantitative relationship between the functional properties of human and macaque V5. Nature Neuroscience, 3, 716–723. Salzman, C. D., Murasugi, C. M., Britten, K. H., & Newsome, W. T. (1992). Microstimulation in visual area MT: Effects on direction discrimination performance. Journal of Neuroscience, 12, 2331–2355. Sclar, G., Maunsell, J. H., & Lennie, P. (1990). Coding of image contrast in central visual pathways in macaque monkey. Vision Research, 30, 1–10. Snowden, R. J., Treue, S., & Andersen, R. A. (1992). The response of neurons in areas V1 and MT of the alert rhesus monkey to moving random dot patterns. Experimental Brain Research, 88, 389–400. Snowden, R. J., Treue, S., Erickson, R. G., & Andersen, R. A. (1991). The response of area MT and V1 neurons to transparent motion. Journal of Neuroscience, 11, 2768–2785. Tootell, R. B., Reppas, J. B., Dale, A. M., Look, R. B., Sereno, M. I., Malach, R., et al. (1995). Visual motion aftereffect in human cortical area MT revealed by functional magnetic resonance imaging. Nature, 375, 139–141. Treue, S., Hol, K., & Rauber, H. J. (2000). Seeing multiple directions of motion – physiology and psychophysics. Nature Neuroscience, 3, 270–275. Treutwein, B. (1995). Adaptive psychophysical procedures. Vision Research, 35, 2503–2522. van Wezel, R. J. A., & Britten, K. H. (2002). Motion adaptation in area MT. Journal of Neurophysiology, 88, 3469–3476. Viswanathan, A., & Freeman, R. D. (2007). Neurometabolic coupling in cerebral cortex reflects synaptic more than spiking activity. Nature Neuroscience, 10, 1308–1312. Watson, J. D., Myers, R., Frackowiak, R. S., Hajnal, J. V., Woods, R. P., Mazziotta, J. C., et al. (1993). Area V5 of the human brain: Evidence from a combined study using positron emission tomography and magnetic resonance imaging. Cerebral Cortex, 3, 79–94. Wiese, M., & Wenderoth, P. (2007). The different mechanisms of the motion direction illusion and aftereffect. Vision Research, 47, 1963–1967.
Contents lists available at ScienceDirect
Vision Research journal homepage: www.elsevier.com/locate/visres
A comparison of monkey and human motion processing mechanisms Catherine Lynn, William Curran ⇑ School of Psychology, Queen’s University Belfast, Belfast, BT7 1NN, UK
a r t i c l e
i n f o
Article history: Received 4 June 2010 Received in revised form 3 August 2010
Keywords: Motion processing Motion adaptation Direction aftereffect Motion-sensitive neurons
a b s t r a c t Single-cell recording studies have provided vision scientists with a detailed understanding of motion processing at the neuronal level in non-human primates. However, despite the development of brain imaging techniques, it is not known to what extent the response characteristics of motion-sensitive neurons in monkey brain mirror those of human motion-sensitive neurons. Using a motion adaptation paradigm, the direction aftereffect, we recently provided evidence of a strong resemblance in the response functions of motion-sensitive neurons in monkey and human to moving dot patterns differing in dot density. Here we describe a series of experiments in which measurements of the direction aftereffect are used to infer the response characteristics of human motion-sensitive neurons when viewing transparent motion and moving patterns that differ in their signal-to-noise ratio (motion coherence). In the case of transparent motion stimuli, our data suggest suppressed activity of motion-sensitive neurons similar to that reported for macaque monkey. In the case of motion coherence, our results are indicative of a linear relationship between signal intensity (coherence) and neural activity; a pattern of activity which also bears a striking similarity to macaque neural activity. These findings strongly suggest that monkey and human motionsensitive neurons exhibit similar response and inhibitory characteristics. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Single-cell recording studies have provided a detailed characterization of the response properties of motion-sensitive neurons in monkey visual cortex (Albright, 1984; Dubner & Zeki, 1971; Krekelberg & Albright, 2005; Lagae, Gulyas, Raiguel, & Orban, 1989; Newsome & Paré, 1988). This approach, which was instrumental in identifying monkey area MT/V5 as an important cortical site for motion processing, typically relies on neuronal spiking as a measure of neural activity. Indeed, spiking discharge of MT/V5 direction-sensitive neurons has been shown to be causally related to behavioural measurements of direction perception (Ditterich, Mazurek, & Shadlen, 2003; Salzman, Murasugi, Britten, & Newsome, 1992). Functional magnetic resonance imaging (fMRI) has been deployed effectively for the study of motion processing in the human brain (Bartels, Logothetis, & Moutoussis, 2008; Morrone et al., 2000; Tootell et al., 1995; Watson et al., 1993), and has been used to infer the response properties of human motion-sensitive neurons. However, how neuronal spiking influences fMRI signal is, as yet, unclear. While some studies have found fMRI signals to be directly proportional to average neuronal spiking rates (Boynton, Demb, Glover, & Heeger, 1999; Rees, Friston, &
⇑ Corresponding author. E-mail address: [email protected] (W. Curran). 0042-6989/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.visres.2010.08.007
Koch, 2000), recent findings suggest that fMRI signals are most likely driven by synaptic activity rather than neuronal spiking activity (Viswanathan & Freeman, 2007). In a review of what can be inferred from fMRI signals, Logothetis (2008) concludes that it is unrealistic at present to draw a direct analogy between neuronal spiking data from animal experiments and fMRI signals derived from human studies. The spike discharge of direction-sensitive neurons in monkey is determined by a number of factors; including how well the stimulus motion direction is matched to the neurons’ preferred direction (Albright, 1984; Snowden, Treue, & Andersen, 1992), and the ‘strength’ of the motion signal. Motion signal strength can be altered by manipulating stimulus contrast (Sclar, Maunsell, & Lennie, 1990), motion coherence (Britten, Shadlen, Newsome, & Movshon, 1993) or dot density (Snowden, Treue, Erickson, & Andersen, 1991; Snowden et al., 1992). When its preferred direction of motion is presented within a neuron’s receptive field for a prolonged period of time a process known as adaptation occurs in which the neuron’s responsiveness is reduced and its direction tuning is modified. Kohn and Movshon (2004) have shown that adaptation of MT direction-sensitive neurons to near-preferred directions causes direction tuning to shift towards the adapted direction. This direction tuning shift is consistent with a perceptual aftereffect known as the direction aftereffect (see below). Stimuli that evoke a strong response in targeted neurons generally act as more effective adaptors (Kohn, 2007; van Wezel & Britten, 2002). More effective motion
2138
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
adaptors, in turn, induce greater changes in neurons’ direction tuning functions, which are thought to impact on the magnitude of motion-adaptation phenomena (Kohn, 2007; Kohn & Movshon, 2004). Given what we know about the response characteristics of monkey direction-sensitive neurons, and how their response functions change following adaptation, it is possible to measure (indirectly) whether human direction-sensitive neurons have similar or different response characteristics to monkey. This can be achieved through the use of motion-adaptation phenomena, such as the direction aftereffect (DAE). The DAE describes the misperception of a motion direction following prolonged viewing of (adaptation to) a different direction of motion (Curran, Clifford, & Benton, 2006a; Levinson & Sekuler, 1976; Patterson & Becker, 1996; Wiese & Wenderoth, 2007). While it has received considerably less attention than the well known motion aftereffect (Mather, Verstraten, & Anstis, 1998), the DAE has contributed to our understanding of speed and direction coding. For example, it has been used to support previous reports of the motion system having at least two speed channels (Curran, Clifford, & Benton, 2006b). Furthermore the DAE and its simultaneous counterpart, direction repulsion (Marshak & Sekuler, 1979; Mather & Moulden, 1980), have been cited as strong evidence for the distribution-shift model of direction coding (Mather & Moulden, 1980); a model which proposes that direction is coded by the population response distribution of motion-sensitive neurons. If motion-sensitive neurons have similar response characteristics in human and monkey, and given the reported relationship between neural spiking and aftereffect magnitude, motion stimuli that are known to have differential effects on the spiking discharge of monkey direction-sensitive neurons should induce DAEs that differ in magnitude. Varying dot density is a prime example of a motion stimulus manipulation which differentially affects neural spiking. Snowden et al. (1991, 1992) reported that macaque motion-sensitive neurons rapidly increase their response to their preferred direction as the number of moving dots within their receptive fields increases, and that this rapid increase in spike discharge plateaus at relatively low dot densities (at approximately 8 dots/deg2). When additional dots moving in the neurons’ anti-preferred direction were added, Snowden et al. (1991) recorded density-tuning functions indicative of a division-like inhibitory interaction between neurons tuned to opposite directions. In a recent study Curran and Lynn (2009) used the DAE to investigate whether human motion-sensitive neurons respond to varying dot density in a similar manner to macaque. It was reasoned that, if the differential responses to dot density found in macaque mirror the response characteristics of human motion-sensitive neurons, low dot density stimuli would be less effective adaptors than high dot density stimuli. Given the reported relationship between neural spiking and aftereffect magnitude (Kohn, 2007; van Wezel & Britten, 2002), the former stimuli should induce a weaker DAE than the latter stimuli. The results were strikingly similar to Snowden et al.’s (1991) single-cell recording data; DAE magnitude increased rapidly with increasing adaptor dot density up to 10 dots/deg2, then levelled off and remained constant for higher dot densities. When the experiment was repeated, but this time with additional dots moving in the opposite direction, the DAE density-tuning functions were consistent with division-like inhibitory interactions between neurons tuned to opposite directions. Furthermore, when equal numbers of dots moved in opposite directions a measurable DAE was still evident; this concurs with Snowden et al.’s (1991) finding that direction-sensitive neurons continue to respond (albeit at a reduced rate) when equal numbers of dots move in the neurons’ preferred and anti-preferred directions. These results strongly suggest that, in the limited case of dot density, response charac-
teristics of macaque motion-sensitive neurons mirror those of human motion-sensitive neurons. The current series of experiments investigates further the extent to which the response characteristics of macaque motionsensitive neurons are an accurate model of human motion processing. Experiment 1 uses DAE magnitude measurements to infer whether the responsiveness of direction-sensitive neurons is suppressed when presented with two transparent motions moving in two distinct directions. Snowden et al. (1991) have previously addressed this question with respect to macaque motion-sensitive neurons. In their experiment they identified the preferred direction for a number of MT neurons; then recorded the neurons’ responses when two superimposed, transparent motions were presented within a neuron’s receptive field. One of the transparent motions moved in the neuron’s preferred direction, and the second motion direction differed from the first by up to 180° (in steps of 45°). Snowden et al. observed suppression, but not complete extinction, of neural spike discharge under the transparent motions conditions. Interestingly, two of the non-preferred directions (45° and 90°) produced an excitatory response when presented in isolation, despite the fact that they produced an inhibitory effect when presented with the preferred direction. Treue, Hol, and Rauber (2000) found similar suppressive effects. These authors recorded the spike discharge of MT neurons as a function of the direction difference between two transparent motions, and as a function of the mean direction of the transparent motions relative to the neuron’s preferred direction. As in the Snowden et al. paper, Treue et al’s data reveal a clear suppressive effect on spike discharge when one of the two transparent motions matched the neuron’s preferred direction. Both Snowden et al., and Treue et al., report that the suppressive effect increases with increasing direction difference between the transparent motions; and that this increasing suppression either levels out (Treue et al., 2000) or slows dramatically (Snowden et al., 1991) for direction differences greater than 90°. If macaque motion-sensitive neuron activity is an accurate model for human motion-sensitive neurons, the macaque data generated by the above studies make clear predictions for DAE measurements in Experiment 1. Two transparent motions that differ in their directions should suppress the responses of neurons preferentially tuned to either of the two directions. Given the relationship between neural spiking and aftereffect magnitude, adapting to such a stimulus should produce a weaker DAE than adapting to one of the directions in isolation. While DAE magnitude will be reduced under transparent motion conditions, the macaque data dictate that it will not be extinguished; but, rather, should level out for transparent adaptors whose component direction difference exceeds 90°. Experiment 2 uses DAE magnitude measurements to infer the relationship between motion coherence strength and neural spiking activity. Motion coherence strength in a random dot display is determined by the proportion of dots that move in a single, coherent direction. In the case of 0% motion coherence, all dots are randomly re-plotted for each frame refresh. In a stimulus with 50% motion coherence, half the dots are randomly replotted for each frame and half are re-plotted with a specific spatiotemporal relation to dots displayed in the previous frame. Britten et al. (1993) recorded the spike discharge of macaque MT neurons as a function of the motion coherence of stimuli presented within the neurons’ receptive fields. They found that, for most of the neurons, spike discharge increased linearly as a function of motion coherence level. If macaque and human motion-sensitive neurons respond similarly to changes in motion coherence, Britten et al’s data predict that DAE magnitude should increase linearly with increasing motion coherence of the adapting stimulus.
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
2. Experiment 1. Responses of human motion-sensitive neurons to transparent motion mirror those of macaque 2.1. Observers Four observers – one of the authors and three naïve participants – took part in the experiment.
2139
vertically upwards. DAE magnitude was taken as the difference between the two directions (perceived and actual). A range of direction differences was used for the transparent adaptor stimuli (30°, 60°, 90° and 120°). Observers generated four psychometric functions each for the baseline and transparency conditions, with each psychometric function being derived from 64 trials. The interval between testing with different adaptors was at least 15 min, thus ensuring recovery from adaptation.
2.2. Apparatus and stimuli 3. Results Stimuli were displayed on a Mitsubishi 2070SB monitor, and comprised random dot kinematograms (RDK) presented within a circular aperture (7 deg2), with each RDK containing equal numbers of black and white dots (dot diameter = 1.8 arcmin) against a mean luminance background (59 cd per m2). The monitor was driven by a Cambridge Research Systems Visage graphics board at a frame rate of 80 Hz. 2.3. Procedure During the initial motion adaptation phase (30 s duration) observers were presented with a random dot adapting stimulus, in which either all dots moved 45° to the right of vertical upwards (baseline condition) or half the dots moved in the baseline direction and remaining dots moved in a direction clockwise from 45° (transparency condition). Adaptor speed was fixed at 2.5° s 1, and adaptor dot density was set to either 28 dots/deg2 (baseline condition) or 56 dots/deg2 (transparency condition). The adaptor direction(s) was the same for all subsequent top-up phases. To help maintain fixation, a central spot was present throughout each experimental run. In the test phase following adaptation, observers judged whether the test stimulus (speed 2.5° s 1, dot density 28 dots/deg2, duration 200 ms) was moving left or right of vertical up. A vertical line extended from the top and bottom of the test stimulus aperture, thus providing observers with a reference to vertical. In order to maintain adaptation, test phases alternated with adaptation top-up phases of 5 s duration. Test stimulus motion direction was chosen by an adaptive method of constant stimuli (adaptive probit estimation), a method that dynamically updates the set of stimulus motion directions being presented depending on the observer’s previous responses (Treutwein, 1995). The stimulus values are selected to optimize the estimation of the point of subjective equality (PSE), in our case the direction the test stimulus was moving when it was perceived as moving
Fig. 1a shows the same pattern of results for all observers. DAE magnitude is greatest for the baseline condition, when all adaptor dots move in the same direction (45°). Under the transparent adaptor conditions, when half the dots move in the baseline direction and half move in a direction clockwise of the baseline direction, there is clear evidence of DAE magnitude decreasing with increasing direction difference between the transparent motion components. This decrease in DAE magnitude continues for direction differences up to 60°, and levels out for larger adaptor direction differences. t-Test analysis reveals a significant reduction in DAE magnitude between the 0° and 30° conditions (df = 15, t = 3.914, p < 0.001), and between the 30° and 60° conditions (df = 15, t = 4.94, p < 0.001). The data suggest that activity of neural populations sensitive to the baseline direction is suppressed when the adaptor stimulus contains an additional direction of motion. This is consistent with macaque data (Snowden et al., 1991), and suggests that human and macaque motion-sensitive neurons respond similarly to transparently moving stimuli. It is interesting to note that, whereas Snowden et al. report a reduction of approximately 70% in neural firing when the preferred and additional motions differ by 90°, DAE magnitude is reduced by approximately 25% under the same condition. This suggests that the relationship between neural activity level and DAE magnitude may not be a straightforward linear one. Snowden et al. (1991) reported that motion directions differing by up to 90° from a neuron’s preferred direction suppressed the neuron’s spike discharge when presented in combination with the preferred motion direction. The suppressive effect developed quickly up to a direction difference of 90° and increased very slowly for larger direction differences. Treue et al.’s (2000) data, on the other hand, reveal a levelling off of the suppressive effect for direction differences greater than 90°. Snowden et al. also found that a subset of suppressive motion directions – up to 90° –
Fig. 1. (a) Direction aftereffect magnitude following adaptation to either a single motion moving in the baseline direction (45° from vertical) or a transparent motion stimulus in which half the dots move in the baseline direction and half move in a direction clockwise from the baseline direction. Light grey symbols plot individual DAE functions for each of the four observers, and the filled circles plot the mean DAE function. Each symbol is the average of four PSEs. Note that the strongest DAE is generated following adaptation to the baseline direction. When dots moving in a different direction are added to the adapting stimulus, DAE magnitude initially decreases with increasing direction difference before levelling out and remaining constant for direction differences of 60° or more. Interestingly, following adaptation to each of the individual directions (b), a DAE is induced by directions 30° and 60° from the baseline direction; yet these same directions have a suppressive effect on the DAE when combined with the baseline direction. Error bars represent ±1 standard error.
2140
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
produced an excitatory response when presented alone. To ascertain whether this finding also applies to human motion processing, we repeated our experiment; but this time had observers adapt to single motion directions (dot density = 28 dots/deg2) ranging from 30° to 120° clockwise from the baseline direction (45°). The results (Fig. 1b) show that adaptors whose directions differed by up to 60° from the baseline direction of Experiment 1 induce a measurable DAE in test stimuli moving vertically up; this is despite the fact that these same directions suppress DAE magnitude when presented with the baseline direction. Statistical analysis shows the effect to be significant (30° from baseline: df = 15, t = 23.86, p < .001; 60° from baseline: df = 15, t = 7.402, p < .001). 4. Experiment 2. Responses of human motion-sensitive neurons to varying motion coherence mirror those of macaque This experiment investigated the effect that varying motion coherence has on DAE magnitude. If human and macaque motion-sensitive neurons respond similarly to motion coherence manipulation, previous single-cell recording data (Britten et al., 1993) would predict a linear relationship between an adapting stimulus’s motion coherence level and DAE magnitude. 4.1. Observers Five observers – one of the authors and four naïve – participated in the experiment. 4.2. Stimuli As in Experiment 1, a random dot display was used for adaptor and test stimuli. The adaptor stimulus’s motion coherence was varied between 0% (all dots randomly re-plotted on each frame) and 100% (all dots moved in the direction 45° right of vertical upwards). For any given motion coherence level, all dots had the same probability of being labelled a signal dot in each frame. The probability that any dot would continue being a signal dot through N frames was cN 1, where c signifies the proportion of coherently moving dots. The test stimulus always contained 100% motion coherence, in which all dots moved in a uniform direction. Adaptor and test stimuli had identical dot densities (64 dots/deg2). 4.3. Procedure Apart from the differences between adaptor stimuli used in Experiments 1 and 2, the procedure was the same as in Experiment 1. Six motion coherence levels were used for the adaptor stimuli – 0%, 20%, 40%, 60%, 80% and 100%. Observers generated four psychometric functions per coherence condition, and took a break of at least 15 min between each adaptor condition to allow recovery from adaptation. 4.4. Results The results from this experiment reveal a strongly linear relationship between adaptor coherence level and DAE magnitude (Fig. 2). As in Experiment 1, these data are strongly consistent with what one would expect if human and macaque motion-sensitive neurons share similar response characteristics. 5. Discussion Much of what we know about the response characteristics of motion-sensitive neurons is derived from single-cell recordings in non-human primates. If one is to develop accurate models of human
Fig. 2. Results from Experiment 2, in which DAE magnitude was measured as a function of the adaptor stimulus’s motion coherence. The data reveal a clear linear relationship between motion coherence level and DAE magnitude.
motion processing, algorithms implemented by such models should be constrained by the known physiological properties of the visual system. To this end, monkey single-cell recording data have played an important role in informing the development of motion models. Implicit in this approach is the assumption that the response characteristics of monkey motion-sensitive neurons mirror faithfully the response characteristics of motion-sensitive neurons in the human brain. The development of brain imaging technology, particularly fMRI, has permitted researchers to gauge the extent to which the response characteristics of human motion-processing neurons mirror those of monkey. Indeed, a number of fMRI studies have revealed striking similarities between human fMRI BOLD signal and monkey single-cell recording data (Huk, Rees, & Heeger, 2001; Krekelberg, Boynton, & van Wezel, 2006; Rees et al., 2000). However, the reported coupling between local field potentials and changes in tissue oxygenation levels (Viswanathan & Freeman, 2007) suggests that it is currently unrealistic to draw a direct analogy between the fMRI BOLD signal from human recordings and neuronal spiking data obtained from monkey experiments (Logothetis, 2008). Thus it is not yet clear to what extent the spiking activity of motion-sensitive neurons in monkey accurately reflects the spiking activity of human motion-sensitive neurons. We used a well known motion-adaptation phenomenon, the direction aftereffect, to infer changes in the spiking activity of motion-sensitive neurons as a function of manipulating transparent motion directions and motion coherence. This indirect approach to inferring neural responsiveness was motivated by the observations that: (1) stimuli that evoke a more robust response in targeted neurons generally act as more effective adaptors (van Wezel & Britten, 2002), and (2) more effective motion adaptors induce greater changes in neurons’ direction tuning functions, which is thought to impact on DAE magnitude (Kohn, 2007; Kohn & Movshon, 2004). By identifying motion stimulus manipulations that are known to result in changes in spiking rate of monkey motion-sensitive neurons, and using these stimuli as adaptors, one can make clear predictions about the impact of these adapting stimuli on DAE magnitude. If DAE magnitude changes are in line with the predictions based on single-cell recording data, this would be compelling evidence that the response characteristics of monkey motion-sensitive neurons mirror those of human neurons. The experiments described here build on recent research suggesting that monkey and human motion-sensitive neurons show the same differential responses to random dot motion stimuli differing in dot density (Curran & Lynn, 2009). The current experiments have extended this comparison between monkey and
C. Lynn, W. Curran / Vision Research 50 (2010) 2137–2141
human to two other types of motion stimuli – transparent and coherent motion. The results from Experiment 1 provide compelling evidence that, just as in macaque (Snowden et al., 1991; Treue et al., 2000), the responsiveness of direction-sensitive neurons to their preferred direction is suppressed when a second motion direction is superimposed on the neurons’ preferred direction. Furthermore, two of the directions that suppressed DAE magnitude were found to induce a measurable DAE when used as a single direction adaptor. This is what would be predicted given Snowden et al’s finding that additional directions suppress the responsiveness of neurons when superimposed on the neurons’ preferred direction, yet have an excitatory effect when presented alone. In the case of motion coherence, the data from Experiment 2 are indicative of a linear increase in neural responsiveness as motion coherence rises. Again, this is in line with monkey single-cell recording data (Britten et al., 1993) in which the majority of MT neurons display spiking activity that increases in a linear fashion with increasing motion coherence. By using the technique of psychophysical adaptation to make inferences about neural processing in human visual cortex, we demonstrate that human motion perception is qualitatively consistent with the operation of the principal animal model of motion processing – macaque area MT. Thus these experimental results support the common held belief that the properties of monkey motion-sensitive neurons can be used as a guide to explain many aspects of human motion processing. References Albright, T. D. (1984). Direction and orientation selectivity of neurons in visual area MT of the macaque. Journal of Neurophysiology, 52, 1106–1130. Bartels, A., Logothetis, N. K., & Moutoussis, K. (2008). FMRI and its interpretations: An illustration on directional selectivity in area V5/MT. Trends in Neuroscience, 31, 444–453. Boynton, G. M., Demb, J. B., Glover, G. H., & Heeger, D. J. (1999). Neuronal basis of contrast discrimination. Vision Research, 39, 257–269. Britten, K. H., Shadlen, M. N., Newsome, W. T., & Movshon, J. A. (1993). Responses of neurons in macaque MT to stochastic motion signals. Visual Neuroscience, 10, 1157–1169. Curran, W., Clifford, C. W., & Benton, C. P. (2006a). The direction aftereffect is driven by adaptation of local motion detectors. Vision Research, 46, 4270–4278. Curran, W., Clifford, C. W. G., & Benton, C. P. (2006b). New binary direction aftereffect does not add up. Journal of Vision, 6, 1451–1458. Curran, W., & Lynn, C. (2009). Monkey and humans exhibit similar motionprocessing mechanisms. Biological Letters, 5, 743–745. Ditterich, J., Mazurek, M. E., & Shadlen, M. N. (2003). Microstimulation of visual cortex affects the speed of perceptual decisions. Nature Neuroscience, 6, 891–898. Dubner, R., & Zeki, S. M. (1971). Response properties and receptive fields of cells in an anatomically defined region of the superior temporal sulcus in the monkey. Brain Research, 35, 528–532. Huk, A. C., Rees, D., & Heeger, D. J. (2001). Neuronal basis of the motion aftereffect reconsidered. Neuron, 32, 161–172.
2141
Kohn, A. (2007). Visual adaptation: Physiology, mechanisms, and functional benefits. Journal of Neurophysiology, 97, 3155–3164. Kohn, A., & Movshon, J. A. (2004). Adaptation changes the direction tuning of macaque MT neurons. Nature Neuroscience, 7, 764–772. Krekelberg, B., & Albright, T. D. (2005). Motion mechanisms in macaque MT. Journal of Neurophysiology, 93, 2908–2921. Krekelberg, B., Boynton, G. M., & van Wezel, R. J. A. (2006). Adaptation: From single cells to BOLD signals. Trends in Neuroscience, 29, 250–256. Lagae, L., Gulyas, B., Raiguel, S., & Orban, G. A. (1989). Laminar analysis of motion information processing in macaque V5. Brain Research, 496, 361–367. Levinson, E., & Sekuler, R. (1976). Adaptation alters perceived direction of motion. Vision Research, 16, 779–781. Logothetis, N. K. (2008). What we can do and what we cannot do with fMRI. Nature, 453, 869–878. Marshak, W., & Sekuler, R. (1979). Mutual repulsion between moving visual targets. Science, 205, 1399–1401. Mather, G., & Moulden, B. (1980). A simultaneous shift in apparent direction: Further evidence for a ‘‘distribution-shift” model of direction coding. Quarterly Journal of Experimental Psychology, 32, 325–333. Mather, G., Verstraten, F., & Anstis, S. (1998). The motion aftereffect. A modern perspective. Cambridge, Massuchusetts: MIT Press. Morrone, M. C., Tosetti, M., Montanaro, D., Fiorentini, A., Cioni, G., & Burr, D. C. (2000). A cortical area that responds specifically to optic flow, revealed by fMRI. Nature Neuroscience, 3, 1322–1328. Newsome, W. T., & Paré, E. B. (1988). A selective impairment of motion perception following lesions of the middle temporal visual area (MT). Journal of Neuroscience, 8, 2201–2211. Patterson, R., & Becker, S. (1996). Direction-selective adaptation and simultaneous contrast induced by stereoscopic (cyclopean) motion. Vision Research, 36, 1773–1781. Rees, G., Friston, K., & Koch, C. (2000). A direct quantitative relationship between the functional properties of human and macaque V5. Nature Neuroscience, 3, 716–723. Salzman, C. D., Murasugi, C. M., Britten, K. H., & Newsome, W. T. (1992). Microstimulation in visual area MT: Effects on direction discrimination performance. Journal of Neuroscience, 12, 2331–2355. Sclar, G., Maunsell, J. H., & Lennie, P. (1990). Coding of image contrast in central visual pathways in macaque monkey. Vision Research, 30, 1–10. Snowden, R. J., Treue, S., & Andersen, R. A. (1992). The response of neurons in areas V1 and MT of the alert rhesus monkey to moving random dot patterns. Experimental Brain Research, 88, 389–400. Snowden, R. J., Treue, S., Erickson, R. G., & Andersen, R. A. (1991). The response of area MT and V1 neurons to transparent motion. Journal of Neuroscience, 11, 2768–2785. Tootell, R. B., Reppas, J. B., Dale, A. M., Look, R. B., Sereno, M. I., Malach, R., et al. (1995). Visual motion aftereffect in human cortical area MT revealed by functional magnetic resonance imaging. Nature, 375, 139–141. Treue, S., Hol, K., & Rauber, H. J. (2000). Seeing multiple directions of motion – physiology and psychophysics. Nature Neuroscience, 3, 270–275. Treutwein, B. (1995). Adaptive psychophysical procedures. Vision Research, 35, 2503–2522. van Wezel, R. J. A., & Britten, K. H. (2002). Motion adaptation in area MT. Journal of Neurophysiology, 88, 3469–3476. Viswanathan, A., & Freeman, R. D. (2007). Neurometabolic coupling in cerebral cortex reflects synaptic more than spiking activity. Nature Neuroscience, 10, 1308–1312. Watson, J. D., Myers, R., Frackowiak, R. S., Hajnal, J. V., Woods, R. P., Mazziotta, J. C., et al. (1993). Area V5 of the human brain: Evidence from a combined study using positron emission tomography and magnetic resonance imaging. Cerebral Cortex, 3, 79–94. Wiese, M., & Wenderoth, P. (2007). The different mechanisms of the motion direction illusion and aftereffect. Vision Research, 47, 1963–1967.