Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation
Saturday, November 09, 2013, 1:00 pm - 3:15 pm
15.08: Causal mapping of lateral prefrontal regions to separate stages of perceptual decision making
*D. RAHNEV, A. LARSON, M. D'ESPOSITO; UC Berkeley, Berkeley, CA
Abstract Body: Perceptual decision-making emerges from the orchestrated activity of large patches of cortex. Recent research has suggested a central role for the prefrontal cortex (PFC) in various stages of visual processing, yet we still lack an overarching theory of PFC’s involvement in perceptual decision-making. Based on previous research, we reasoned that visual processing can be separated into four inter-related stages: (1) predictive, preparatory stage in anticipation of visual input, (2) low-level processing stage, (3) high-level “decisional” stage, and (4) monitoring stage where the whole process is evaluated. Further, we proposed that the PFC is involved in all but the second of the above stages in a specific temporo-spatial manner. More specifically, we hypothesized that the frontal eye fields (FEF) send predictive signals to the visual cortex (stage 1), the dorsolateral prefrontal cortex (DLPFC) acts as a decision stage at the top of visual processing hierarchy (stage 3), and the anterior frontal cortex (aPFC) monitors this process and decides on the reliability of the perceptual decision (stage 4). Here we tested this proposal by applying, on different days, theta-burst transcranial magnetic stimulation (TBS) to each of these three regions (somatosensory cortex, a control region, was stimulated on a fourth day; order was completely randomized). The regions were identified for each individual (N = 18) on a separate day while subjects were performing our psychophysical task. The task combined spatial attention, speed-accuracy manipulation, and confidence ratings. We found evidence for differential involvement of each of the three PFC regions in our task. First, TBS to FEF led to a performance interaction between the side to which subjects were cued to attend and the side where the relevant stimulus appeared. This suggests that attentional mechanisms were affected (stage 1). Second, TBS to DLPFC affected the decisional stage in that it led to decreased ability to follow speed/accuracy instructions (stage 3). Indeed a diffusion model fit to the data showed that after DLPFC stimulation subjects failed to adjust the decision threshold as much as in the control condition in trials in which they were asked to emphasize speed or accuracy. Third, TBS to aPFC led to a change in subjects’ metacognitive ability. That is, it altered the extent to which subjects’ confidence ratings predicted their accuracy on a trial-by-trail basis (stage 4). Overall, our study provides causal evidence for a triple dissociation of three separate stages of perceptual decision-making in the lateral prefrontal cortex.
Lay Language Summary: Seeing the world around us depends on a number of processes that act in synchrony to produce our visual experience. The current research identifies three of these processes and maps them to different parts of the prefrontal cortex in the brain. These findings shed light on the processes that allow us to visually experience the world. The brain can be divided into several cortexes. The visual cortex - located in the back of the head - is where most of the visual processing takes place. On the other hand, the prefrontal cortex - located in the front of the head - is responsible for high-level processes such as thinking and decision-making. Previous research has shown that the prefrontal cortex is involved in visual processing by sending “top-down” signals. However, this is the first study to simultaneously map three different regions of the prefrontal cortex to specific processes related to visual processing. Visual processing can be roughly separated into four different stages: (1) predictive, preparatory stage in anticipation of visual input, (2) low-level processing stage, (3) high-level “decisional” stage, and (4) monitoring stage where the whole process is evaluated. For example, if we are visiting a friend at her new home and we are offered a brief glimpse of the kitchen, we may already expect to see a fridge, microwave, and an oven (stage 1). This expectation will then guide our vision and help us identify these objects. When we take our first look at the kitchen, the information from our eyes is processed so that low-level features such as color and contrast are extracted (stage 2). These features are then combined to form our percepts of objects such as a fridge or an oven even if we did not have the time to take a good look at each of these objects (stage 3). Finally, the process is evaluated and we can later tell another friend about the objects in the kitchen and judge to what extent we had actually noticed the different details about each object (stage 4). The current research tested the involvement of prefrontal cortex in these various stages of visual processing. We delivered theta-burst transcranial magnetic stimulation (TMS) to human participants just before they performed a visual task. TMS has been shown to transiently suppress the normal working of the targeted part of the brain. The effect lasts for up to an hour and has no known side effects beyond the time of stimulation. We delivered TMS to three different areas of the right prefrontal cortex that lie on an axis from the border of the forehead to the top of the head just above the ear. The regions were identified for each participant on a separate day in which subjects performed a visual task while their brain activity was monitored in an MRI scanner. Our results showed that delivering TMS to the three different parts of the prefrontal cortex disrupts different stages of the visual processing. Targeting the areas that lie from front to back disrupted the evaluative (stage 4), decisional (stage 3), and preparatory processes (stage 1), respectively. We have the impression that our vision is like a video camera: the world just happens in front of us and we capture it as it is. This study demonstrates that our vision is in fact fractured into several processes that depend on activity in separate brain areas not only in visual cortex but also in the prefrontal cortex, suggesting that perception is a process of active inference rather than passive video-camera-like recording.
Neuroscience 2013 (43rd annual meeting of the Society for Neuroscience)Exit