Research

Peer M*, Nadar C*, Epstein RA, "The format of the cognitive map depends on the structure of the environment", Journal of Experimental Psychology: General 2023 (* equal contribution)

In this study, we studied how features of the environment affect how it is represented in the brain. We reasoned that open environments with clear visibility and surrounding orienting landmarks (e.g. a city park) could be easily represented using a mental Euclidean reference frame. In contrast, environments with diverging paths where visibility is limited (e.g. the inside of a building) could be more easily represented as a cognitive graphs of the paths leading from one location to another, without having to constantly track our global orientation and location. To test this, we built several virtual environments with a similar structure (9 rooms connected by pathways) but with varying of these features (e.g. an open park with dirt roots vs. a maze with walled corridors). Subsequent behavioral tests confirmed that people form a graph-like representation of the closed maze environment vs. a Euclidean representation of more open environments. These findings highlight the importance of environmental features in the formation of different types of cognitive maps, and caution against generalizing across findings from experiments that have used different types of environments.

Peer M, Epstein RA, "The human brain uses spatial schemas to represent segmented environments", Current Biology 2021

In this study, we investigated how the human brain represents environments that are separated into segments. We had subjects navigate in a virtual environment - a park divided to two parts by a river; we then scanned the subjects using fMRI and identified the relations between patterns arising as they imagined different park locations. We found that the occipital place area (OPA) and hippocampus encoded local maps of the two environmental segments, and these maps reflected the common geometric structure of the two subspaces (that is, they used the same spatial schema). This suggests that spatial schemas might be a key component of cognitive maps, and that the brain uses local codes to represent segmented environments instead of a global unified map that extends across the whole environment.

Peer M*, Brunec I*, Newcombe N, Epstein RA, "Structuring Knowledge with Cognitive Maps and Cognitive Graphs", TICS 2021  (* equal contribution)

In this review, we discussed the issue of how cognitive maps are structured, in space and in other domains. While the classical view is that cognitive maps take the form of a Euclidean coordinate system on which locations are embedded using their exact coordinates, other theories suggested that these maps are stored instead as cognitive graphs indicating the routes that connect between locations. We discuss evidence for both types of theories, and suggest that the brain may alternatively use either graphs or Euclidean maps to encode information, and these may rely on partially-separate brain systems. We further suggest that graphs and maps are general frameworks for knowledge organization that can be used to represent different types of information outside the spatial domain (e.g. a social network of how people are connected can be encoded using a cognitive graph, while the difference between people's personalities can be encoded using a Euclidean space representing personality dimensions).

Peer M*, Salomon R*, Goldberg I, Blanke O, Arzy S, "Brain system for mental orientation in space, time, and person", PNAS 2015 (* equal contribution)

in this study, we investigated whether the same brain systems are active for processing of relations in different cognitive domains – space, time and the social domain. Subjects made distance comparisons between familiar places (spatial distance), between events in time (temporal distance), and between people in the social domain (social / emotional distance), under high-resolution 7-Tesla fMRI scanning. We found a common system processing all three domains, overlapping with the brain’s default-mode network, with subdivision to regions active when processing relations in different cognitive domains. These findings suggest that the brain processes relations in multiple cognitive domains using the same brain system, that  has specialized subregions for each domain.

Peer M, Ron Y, Monsa R, Arzy S, "Processing of different spatial scales in the human brain", eLife 2019

In this project we asked whether the same brain systems are active for processing of space at different spatial scales. Subjects performed spatial judgments between familiar locations in environments of different scales – room, building, neighborhood, city, country, and continent. Surprisingly, scale selectivity across the brain was organized in a gradients, where in each spatially-sensitive region the posterior areas were most active for the smallest scales (room, building, neighborhood) and the anterior regions for larger scales (city, country, continent). The posterior parts of the gradients corresponded to known spatial processing regions – PPA, RSC and OPA. Our findings suggest that the brain’s spatial processing system extends further along the cortex than previously suggested, and is organized according to spatial scale.

Peer M, Hayman M, Tamir B, Arzy S, "Brain coding of social network structure", Journal of Neuroscience 2021

In this project we studied another cognitive domain and its organization – the social domain. Social networks are complex, and in order to navigate our social lives we need to know the relations between people in them– but it is unclear how the brain represents and manages knowledge of social relations. To look at this question we used subjects’ Facebook accounts to map all of the relations between people they know, and also asked them to answer questions about these people’s personality, appearance and affiliation to themselves, under fMRI. Using representational similarity analysis of multivoxel activity patterns, we found that different subparts of the default-mode system represent knowledge about how other people are related to each other, how they are related to ourselves, and how similar they are to each other. Furthermore, this coding was prominently found in regions related to spatial processing – suggesting a shared mechanism for knowledge representation across domains.

Peer M, Lyon R, Arzy S, "Orientation and disorientation: Lessons from patients with epilepsy", Epilepsy & Behavior 2014

We reviewed the literature related to disorientation disorders – their different types and the underlying neural systems. Although disorientation in space, time and person is a common phenomenon across different neurological and psychiatric disorders, its underlying brain basis remains unclear. We looked at the different disorders of disorientation in the literature and interviewed subjects suffering from disorientation, and suggested that disorientation can be organized according to two axes – domain (space, time or person) and scale (immediate surroundings vs. the larger world).

Peters-Founshtein G, Peer M, Rein Y, Kahana Merhavi S, Meiner Z, Arzy S, "Mental-orientation: A new approach to assessing patients across the Alzheimer’s disease spectrum.", Neuropsychology 2018

In this project, we attempted to use our previous findings to develop a practical test for Alzheimer’s disease based on disorientation in space, time and person. We developed an in-depth orientation test and found that it performs better than common tests for testing early Alzheimer’s disease, as this disease targets the exact same regions identified as responsible for orientation in space, time and the social domain. The disorientation test is now being developed as a computerized screening tool for clinical practice.

Peer M, Nitzan M, Goldberg I, Katz J, Gomori JM, Ben-Hur T, Arzy S, "Reversible functional connectivity disturbances during transient global amnesia", Annals of Neurology 2014

In this study we used resting-state fMRI to identify the brain basis for transient global amnesia (TGA). Subjects with TGA suffer from complete inability to remember new information (anterograde amnesia), that appears in an sudden manner, and is gradually resolved after several hours. We scanned these subjects during and after the amnesic stage, and identified a widespread functional connectivity disturbance centered on their medial temporal lobe, that gradually disappeared as the subjects recovered from their amnesia.

Peer M, Pruss H, Ben-Dayan I, Paul F, Arzy S, Finke C, "Functional connectivity of large-scale brain networks in patients with anti-NMDA receptor encephalitis: an observational study", The Lancet Psychiatry 2017

In this project we used resting-state fMRI to investigate the brain systems implicated in anti-NMDA receptor encephalitis – an autoimmune disorder that manifests as an acute psychotic state with additional cognitive symptoms. We found that patients had a major disturbance in fronto-parietal and fronto-temporal connectivity, which was related to their disease symptoms – fronto-parietal connectivity impairment was correlated with psychiatric symptoms, while fronto-temporal connectivity impairment was correlated with memory symptoms.

Monsa R*, Peer M*, Arzy S, "Self‐reference, emotion inhibition and somatosensory disturbance: preliminary investigation of network perturbations in conversion disorder", European Journal of Neurology 2018 (*Equal contribution)

In this study we looked at patients with conversion disorder – body paralysis and reduced sensation, resulting from psychological stress and not from an organic brain dysfunction. We identified connectivity changes inside the default-mode network and in sensory-motor regions, providing a physiological brain correlate to this intriguing psychiatric disorder.

Peer M, Nitzan M, Bick A, Levin N, Arzy S, "Evidence for functional networks within the human brain's white matter", Journal of Neuroscience 2017

Here we attempted to expand the use of functional MRI to investigations of the brain’s white matter. Although fMRI signal changes are traditionally thought to represent only activity in neuron bodies, recent evidence points to fMRI signals with meaningful functional information along white matter tracts. We used resting-state fMRI to identify clusters of correlated activity changes within the white matter, which were replicated across several datasets of hundreds of subjects, and corresponded to combinations of white-matter tracts. Our findings demonstrate an internal functional organization of the white matter and expand the utility of fMRI to white matter investigations.

Peer M*, Abboud S*, Hertz U, Amedi A, Arzy S, "Intensity‐based masking: A tool to improve functional connectivity results of resting‐state fMRI", Human Brain Mapping 2016 (* Equal contribution)

In this study we investigated the effects of susceptibility artifacts on resting-state functional MRI measures. Many studies use anatomical atlases to investigate connectivity between regions, but neglect to take into account the effect of susceptibility artifacts, which add noise to the signal from many of these regions. We demonstrated the extent of this effect, and suggested a method – intensity-based masking – to reduce these effects and recover true functional information.