Functional mechanisms of probabilistic inference in feature- and space-based attentional systems

Segregation of Neural Circuits Involved in Social Gaze and Non-Social Arrow Cues: Evidence from an Activation Likelihood Estimation Meta-Analysis

Orienting attention by social gaze cues shares some characteristics with orienting attention by non-social arrow cues, but it is unclear whether they rely on similar neural mechanisms. The present ALE-meta-analysis assessed the pattern of brain activation reported in 40 single experiments (18 with arrows, 22 with gaze), with a total number of 806 participants. Our findings show that the network for orienting attention by social gaze and by non-social arrow cues is in part functionally segregated. Orienting by both types of cues relies on the activity of brain regions involved in endogenous attention (the superior frontal gyrus). Importantly, only orienting by gaze cues was also associated with the activity of brain regions involved in exogenous attention (medial frontal gyrus), processing gaze, and mental state attribution (superior temporal sulcus, temporoparietal junction).

Superior Attentional Efficiency of Auditory Cue via the Ventral Auditory-thalamic Pathway

Auditory commands are often executed more efficiently than visual commands. However, empirical evidence on the underlying behavioral and neural mechanisms remains scarce. In two experiments, we manipulated the delivery modality of informative cues and the prediction violation effect and found consistently enhanced RT benefits for the matched auditory cues compared with the matched visual cues. At the neural level, when the bottom-up perceptual input matched the prior prediction induced by the auditory cue, the auditory-thalamic pathway was significantly activated. Moreover, the stronger the auditory-thalamic connectivity, the higher the behavioral benefits of the matched auditory cue. When the bottom-up input violated the prior prediction induced by the auditory cue, the ventral auditory pathway was specifically involved. Moreover, the stronger the ventral auditory-prefrontal connectivity, the larger the behavioral costs caused by the violation of the auditory cue. In addition, the dorsal frontoparietal network showed a supramodal function in reacting to the violation of informative cues irrespective of the delivery modality of the cue. Taken together, the results reveal novel behavioral and neural evidence that the superior efficiency of the auditory cue is twofold: The auditory-thalamic pathway is associated with improvements in task performance when the bottom-up input matches the auditory cue, whereas the ventral auditory-prefrontal pathway is involved when the auditory cue is violated.

Simultaneous modeling of reaction times and brain dynamics in a spatial cueing task

Understanding how brain activity translates into behavior is a grand challenge in neuroscientific research. Simultaneous computational modeling of both measures offers to address this question. The extension of the dynamic causal modeling (DCM) framework for blood oxygenation level‐dependent (BOLD) responses to behavior (bDCM) constitutes such a modeling approach. However, only very few studies have employed and evaluated bDCM, and its application has been restricted to binary behavioral responses, limiting more general statements about its validity. This study used bDCM to model reaction times in a spatial attention task, which involved two separate runs with either horizontal or vertical stimulus configurations. We recorded fMRI data and reaction times (n= 26) and compared bDCM with classical DCM and a behavioral Rescorla–Wagner model using Bayesian model selection and goodness of fit statistics. Results indicate that bDCM performed equally well as classical DCM when modeling BOLD responses and as good as the Rescorla–Wagner model when modeling reaction times. Although our data revealed practical limitations of the current bDCM approach that warrant further investigation, we conclude that bDCM constitutes a promising method for investigating the link between brain activity and behavior.

Large-scale brain networks underlying non-spatial attention updating: Towards understanding the function of the temporoparietal junction

The temporoparietal junction (TPJ) and related areas are activated when a target stimulus appears at unexpected locations in Posner's spatial-cueing paradigm, and also when deviant stimuli are presented within a series of standard events in oddball paradigms. This type of activation corresponds to the ventral attention network (VAN), for regions defined on the basis of the spatial task. However, involvement of the VAN in object-based updating of attention has rarely been examined. In the present study, we used functional magnetic resonance imaging to investigate brain responses to (i) invalid targets after category-cueing and (ii) neutrally cued targets deviating in category from the background series of pictures. Bilateral TPJ activation was observed in response to invalidly cued targets, as compared to neutrally cued targets. Reference to the main large-scale brain networks showed that peaks of this activation located in the angular gyrus and inferior parietal lobule belonged to the default mode (DMN) and fronto-parietal networks (FPN), respectively. We found that VAN regions were involved only for simple detection activity. We conclude that spatial and non-spatial reorienting of attention rely on different network underpinnings. Our data suggest that DMN and FPN activity may support the ability to disengage from contextually irrelevant information.

Lateralization, functional specialization, and dysfunction of attentional networks

The present review covers the latest findings on the lateralization of the dorsal and ventral attention systems, their functional specialization, and their clinical relevance for stroke-induced attentional dysfunction. First, the original assumption of a bilateral dorsal system for top-down attention and a right-lateralized ventral system for stimulus-driven attention is critically reviewed. The evidence for the involvement of the left parietal cortex in attentional functions is discussed and findings on putative pathways linking the dorsal and ventral network are presented. In the second part of the review, we focus on the different attentional subsystems and their lateralization, discussing the differences between spatial, feature- and object-based attention, and motor attention. We also review studies based on predictive coding frameworks of attentional functions. Finally, in the third section, we provide an overview of the consequences of specific disruption within the attention networks after stroke. The role of the interhemispheric (im)balance is discussed, and the results of new promising therapeutic approaches employing brain stimulation techniques such as transcranial magnetic stimulation (TMS) or transcranial direct current stimulation (tDCS) are presented.

Combined expectancies: the role of expectations for the coding of salient bottom-up signals

The visual system forms predictions about upcoming visual features based on previous visual experiences. Such predictions impact on current perception, so that expected stimuli can be detected faster and with higher accuracy. A key question is how these predictions are formed and on which levels of processing they arise. Particularly, predictions could be formed on early levels of processing, where visual features are represented separately, or might require higher levels of processing, with predictions formed based on full object representations that involve combinations of visual features. In four experiments, the present study investigated whether the visual system forms joint prediction errors or whether expectations about different visual features such as color and orientation are formed independently. The first experiment revealed that task-irrelevant and implicitly learned expectations were formed independently when the features were separately bound to different objects. In a second experiment, no evidence for a mutual influence of both types of task-irrelevant and implicitly formed feature expectations was observed, although both visual features were assigned to the same objects. A third experiment confirmed the findings of the previous experiments for explicitly rather than implicitly formed expectations. Finally, no evidence for a mutual influence of different feature expectations was observed when features were assigned to a single centrally presented object. Overall, the present results do not support the view that object feature binding generates joint feature-based expectancies of different object features. Rather, the results suggest that expectations for color and orientation are processed and resolved independently at the feature level.

Parcellation-based tractographic modeling of the dorsal attention network

INTRODUCTION

The dorsal attention network (DAN) is an important mediator of goal-directed attentional processing. Multiple cortical areas, such as the frontal eye fields, intraparietal sulcus, superior parietal lobule, and visual cortex, have been linked in this processing. However, knowledge of network connectivity has been devoid of structural specificity.

METHODS

Using attention-related task-based fMRI studies, an anatomic likelihood estimation (ALE) of the DAN was generated. Regions of interest corresponding to the cortical parcellation scheme previously published under the Human Connectome Project were co-registered onto the ALE in MNI coordinate space and visually assessed for inclusion in the network. DSI-based fiber tractography was performed to determine the structural connections between relevant cortical areas comprising the network.

RESULTS

Twelve cortical regions were found to be part of the DAN: 6a, 7AM, 7PC, AIP, FEF, LIPd, LIPv, MST, MT, PH, V4t, VIP. All regions demonstrated consistent u-shaped interconnections between adjacent parcellations. The superior longitudinal fasciculus connects the frontal, parietal, and occipital areas of the network.

CONCLUSIONS

We present a tractographic model of the DAN. This model comprises parcellations within the frontal, parietal, and occipital cortices principally linked through the superior longitudinal fasciculus. Future studies may refine this model with the ultimate goal of clinical application.

Age-related changes in Bayesian belief updating during attentional deployment and motor intention

Predicting upcoming events using past observations is a crucial component of an efficient allocation of attentional resources. Therefore, the deployment of attention is sensitive to different types of cues predicting upcoming events. Here we investigated probabilistic inference abilities in spatial and feature-based attentional, as well as in motor-intentional subsystems, focusing specifically on the age-related changes in these abilities. In two behavioral experiments, younger and older adults (20 younger and 20 older adults for each experiment) performed three versions of a cueing paradigm, where spatial, feature, or motor cues predicted the location, color, or motor response of a target stimulus. The percentage of cue validity (i.e., the probability of the cue being valid) changed over time, thereby creating a volatile environment. A Bayesian hierarchical model was used to estimate trial-wise beliefs concerning the cue validity from reaction times and to derive a subject-specific belief updating parameter ω in each task version. We also manipulated task difficulty: participants performed an easier version of the task in Experiment 1 and a more difficult version in Experiment 2. Results from Experiment 1 suggested a preserved ability of older adults to use the three different cues to generate predictions. However, the increased task demands of Experiment 2 uncovered a difference in belief updating between the two age groups, indicating moderate evidence for a reduction of the ability to update predictions with motor intention cues in older adults. These results point at a distinction of attentional and motor-intentional subsystems, with age-related differences tackling especially the motor-intentional subsystem.

Integrating modality-specific expectancies for the deployment of spatial attention

The deployment of spatial attention is highly sensitive to stimulus predictability. Despite evidence for strong crossmodal links in spatial attentional systems, it remains to be elucidated how concurrent but divergent predictions for targets in different sensory modalities are integrated. In a series of behavioral studies, we investigated the processing of modality-specific expectancies using a multimodal cueing paradigm in which auditory cues predicted the location of visual or tactile targets with modality-specific cue predictability. The cue predictability for visual and tactile targets was manipulated independently. A Bayesian ideal observer model with a weighting factor was applied to trial-wise individual response speed to investigate how the two probabilistic contexts are integrated. Results showed that the degree of integration depended on the level of predictability and on the divergence of the modality-specific probabilistic contexts (Experiments 1–2). However, when the two probabilistic contexts were matched in their level of predictability and were highly divergent (Experiment 3), higher separate processing was favored, especially when visual targets were processed. These findings suggest that modality-specific predictions are flexibly integrated according to their reliability, supporting the hypothesis of separate modality-specific attentional systems that are however linked to guarantee an efficient deployment of spatial attention across the senses.

Understanding active sampling strategies: Empirical approaches and implications for attention and decision research

In natural behavior we actively gather information using attention and active sensing behaviors (such as shifts of gaze) to sample relevant cues. However, while attention and decision making are naturally coordinated, in the laboratory they have been dissociated. Attention is studied independently of the actions it serves. Conversely, decision theories make the simplifying assumption that the relevant information is given, and do not attempt to describe how the decision maker may learn and implement active sampling policies. In this paper I review recent studies that address questions of attentional learning, cue validity and information seeking in humans and non-human primates. These studies suggest that learning a sampling policy involves large scale interactions between networks of attention and valuation, which implement these policies based on reward maximization, uncertainty reduction and the intrinsic utility of cognitive states. I discuss the importance of using such paradigms for formalizing the role of attention, as well as devising more realistic theories of decision making that capture a broader range of empirical observations.

Disruption of the Right Temporoparietal Junction Impairs Probabilistic Belief Updating

Generating and updating probabilistic models of the environment is a fundamental modus operandi of the human brain. Although crucial for various cognitive functions, the neural mechanisms of these inference processes remain to be elucidated. Here, we show the causal involvement of the right temporoparietal junction (rTPJ) in updating probabilistic beliefs and we provide new insights into the chronometry of the process by combining online transcranial magnetic stimulation (TMS) with computational modeling of behavioral responses. Female and male participants performed a modified location-cueing paradigm, where false information about the percentage of cue validity (%CV) was provided in half of the experimental blocks to prompt updating of prior expectations. Online double-pulse TMS over rTPJ 300 ms (but not 50 ms) after target appearance selectively decreased participants' updating of false prior beliefs concerning %CV, reflected in a decreased learning rate of a Rescorla-Wagner model. Online TMS over rTPJ also impacted on participants' explicit beliefs, causing them to overestimate %CV. These results confirm the involvement of rTPJ in updating of probabilistic beliefs, thereby advancing our understanding of this area's function during cognitive processing.SIGNIFICANCE STATEMENT Contemporary views propose that the brain maintains probabilistic models of the world to minimize surprise about sensory inputs. Here, we provide evidence that the right temporoparietal junction (rTPJ) is causally involved in this process. Because neuroimaging has suggested that rTPJ is implicated in divergent cognitive domains, the demonstration of an involvement in updating internal models provides a novel unifying explanation for these findings. We used computational modeling to characterize how participants change their beliefs after new observations. By interfering with rTPJ activity through online transcranial magnetic stimulation, we showed that participants were less able to update prior beliefs with TMS delivered at 300 ms after target onset.

Spatial Attention, Motor Intention, and Bayesian Cue Predictability in the Human Brain

Predictions about upcoming events influence how we perceive and respond to our environment. There is increasing evidence that predictions may be generated based upon previous observations following Bayesian principles, but little is known about the underlying cortical mechanisms and their specificity for different cognitive subsystems. The present study aimed at identifying common and distinct neural signatures of predictive processing in the spatial attentional and motor intentional system. Twenty-three female and male healthy human volunteers performed two probabilistic cueing tasks with either spatial or motor cues while lying in the fMRI scanner. In these tasks, the percentage of cue validity changed unpredictably over time. Trialwise estimates of cue predictability were derived from a Bayesian observer model of behavioral responses. These estimates were included as parametric regressors for analyzing the BOLD time series. Parametric effects of cue predictability in valid and invalid trials were considered to reflect belief updating by precision-weighted prediction errors. The brain areas exhibiting predictability-dependent effects dissociated between the spatial attention and motor intention task, with the right temporoparietal cortex being involved during spatial attention and the left angular gyrus and anterior cingulate cortex during motor intention. Connectivity analyses revealed that all three areas showed predictability-dependent coupling with the right hippocampus. These results suggest that precision-weighted prediction errors of stimulus locations and motor responses are encoded in distinct brain regions, but that crosstalk with the hippocampus may be necessary to integrate new trialwise outcomes in both cognitive systems.SIGNIFICANCE STATEMENT The brain is able to infer the environments' statistical structure and responds strongly to expectancy violations. In the spatial attentional domain, it has been shown that parts of the attentional networks are sensitive to the predictability of stimuli. It remains unknown, however, whether these effects are ubiquitous or if they are specific for different cognitive systems. The present study compared the influence of model-derived cue predictability on brain activity in the spatial attentional and motor intentional system. We identified areas with distinct predictability-dependent activation for spatial attention and motor intention, but also common connectivity changes of these regions with the hippocampus. These findings provide novel insights into the generality and specificity of predictive processing signatures in the human brain.