Perceptual difficulty modulates the direction of information flow in familiar face recognition

Karimi-Rouzbahani, Hamid; Ramezani, Farzad; Woolgar, Alexandra; Rich, Anina N.; Ghodrati, Masoud

doi:10.1016/j.neuroimage.2021.117896

Cited by 42 publications

(57 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In sum, we propose a distinction between easy tasks, processed mainly in individual brain areas such as the fusiform face area (FFA) and the OFA, and difficult tasks, processed by the involvement of a more widespread network. Although there is little evidence to date directly supporting this proposition (e.g., but see Dowdle et al 2021 ; Karimi-Rouzbahani et al 2021 ), it is compatible with the fMRI literature on neural efficiency. In this literature, functional brain activation of individuals with high and low cognitive abilities has been measured while participants worked on cognitive tasks ( Neubauer and Fink 2009 ).…”

Section: Neurocognitive Mechanisms Underlying Specificity In Accuracy and Speedsupporting

confidence: 77%

Reflections and New Perspectives on Face Cognition as a Specific Socio-Cognitive Ability

2021

View full text Add to dashboard Cite

The study of socio-cognitive abilities emerged from intelligence research, and their specificity remains controversial until today. In recent years, the psychometric structure of face cognition (FC)—a basic facet of socio-cognitive abilities—was extensively studied. In this review, we summarize and discuss the divergent psychometric structures of FC in easy and difficult tasks. While accuracy in difficult tasks was consistently shown to be face-specific, the evidence for easy tasks was inconsistent. The structure of response speed in easy tasks was mostly—but not always—unitary across object categories, including faces. Here, we compare studies to identify characteristics leading to face specificity in easy tasks. The following pattern emerges: in easy tasks, face specificity is found when modeling speed in a single task; however, when modeling speed across multiple, different easy tasks, only a unitary factor structure is reported. In difficult tasks, however, face specificity occurs in both single task approaches and task batteries. This suggests different cognitive mechanisms behind face specificity in easy and difficult tasks. In easy tasks, face specificity relies on isolated cognitive sub-processes such as face identity recognition. In difficult tasks, face-specific and task-independent cognitive processes are employed. We propose a descriptive model and argue for FC to be integrated into common taxonomies of intelligence.

show abstract

Section: Neurocognitive Mechanisms Underlying Specificity In Accuracy and Speedsupporting

confidence: 77%

Reflections and New Perspectives on Face Cognition as a Specific Socio-Cognitive Ability

2021

View full text Add to dashboard Cite

show abstract

“…However, the predictive power of a feature about behavior might not be as important for BCI where the goal is to maximize the accuracy of the commands sent to a computer or an actuator. In contrast, one of the most critical questions in cognitive neuroscience to understand whether the neural signatures that we observe are meaningful in bringing about behavior, as opposed to being epiphenomenal to our experimental setup (e.g., Williams et al, 2007;Jacobs et al, 2009;Ritchie et al, 2015;Hebart and Baker;Woolgar et al, 2019;Karimi-Rouzbahani et al, 2021a;Karimi-Rouzbahani et al, 2021b). To address this point, we evaluated whether our extracted features and their combinations were behaviorally relevant, by correlating our decoding patterns with the behavioral object recognition performance (reaction times in Dataset 2).…”

Section: Discussionmentioning

confidence: 99%

“…UDFS, lasso and ufsol were the best performing FS algorithms leading to the highest maximum and average decoding accuracies (Supplementary Figure 1A and B; black boxes). Dataset 2 tended to yield higher decoding accuracies compared to the other datasets, which might be attributed to the longer presentation time of the stimuli and the active task of the participants, (Karimi-Rouzbahani et al, 2021a;Karimi-Rouzbahani et al, 2021c;Roth et al, 2020). UDFS, ufsol and relief were among the earliest FS algorithms to reach their first above-chance and maximum decoding accuracies (Supplementary Figure 1C and D).…”

Section: Do Different Ways Of Combining Individual Features Affect the Level And Temporal Dynamics Of Information Decoding?mentioning

confidence: 97%

“…Critically, as opposed to the Brain-Computer Interface (BCI) community, where the goal of feature extraction is to maximize the decoding accuracy, in cognitive neuroscience the goal is to find better neural correlates for the behavioral effect under study (Hebart and Baker, 2018;Williams et al, 2007;Jacobs et al, 2009;Woolgar et al, 2019;Karimi-Rouzbahani et al, 2021a;Karimi-Rouzbahani et al, 2021b). Specifically, a given feature is arguably only informative if it predicts behavior.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

When the whole is less than the sum of its parts: maximum object category information and behavioral prediction in multiscale activation patterns

Karimi-Rouzbahani

Woolgar

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Neural codes are reflected in complex, temporally and spatially specific patterns of activation. One popular approach to decode neural codes in electroencephalography (EEG) is multivariate decoding. This approach examines the discriminability of activity patterns across experimental conditions to test if EEG contains information about those conditions. However, conventional decoding analyses ignore aspects of neural activity which are informative. Specifically, EEG data can be decomposed into a large number of mathematically distinct features (e.g., entropy, Fourier and Wavelet coefficients) which can reflect different aspects of neural activity. We previously compared 30 such features of EEG data, and found that visual category, and participant behavior, can be more accurately predicted using multiscale spatiotemporally sensitive Wavelet coefficients than mean amplitude (Karimi-Rouzbahani et al., 2021b). Here, we considered that even this larger set of features may only partially capture the underlying neural code, because the brain could use a combination of encoding protocols within a single trial which is not reflected in any one mathematical feature alone. To check, we combined those mathematical features using state-of-the-art supervised and unsupervised feature selection procedures (n = 17). Across 3 datasets, we compared decoding of visual object category between these 17 sets of combined features, and between combined and individual features. Object category could be robustly decoded using the combined features from all of the 17 algorithms. However, the combination of features, which were equalized in dimension to the individual features, were outperformed in most of the time points by the most informative individual feature (Wavelet coefficients). Moreover, the Wavelet coefficients also explained the behavioral performance more accurately than the combined features. These results suggest that a single but multiscale encoding protocol may capture the neural code better than any combination of features. Our findings put new constraints on the models of neural information encoding in EEG.

show abstract

“…For example, for heterogeneous ROIs, where multiple response modes co-exist, projecting multivariate response patterns onto one dimension could lead to strong distortions. This has led to a recent shift from univariate to multidimensional (multivariate) connectivity analyses (Coutanche and Thompson-Schill, 2013; Goddard et al, 2016; Anzellotti and Coutanche, 2018; Basti et al, 2019; Basti et al, 2020; Karimi-Rouzbahani et al, 2021a; Karimi-Rouzbahani et al, 2021c; Shahbazi et al, 2021). One approach to multidimensional connectivity analyses is Representational Connectivity Analysis (RCA; Kriegeskorte et al, 2008), which utilizes the versatility of Representational Similarity Analysis (RSA) to move from the direct comparison of representations to the comparison of representational geometries (Kriegeskorte et al, 2008).…”

Section: Introductionmentioning

confidence: 99%

Caveats and nuances of model-based and model-free representational connectivity analysis

Karimi-Rouzbahani

Woolgar

Henson

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Brain connectivity analyses have conventionally relied on statistical relationship between one-dimensional summaries of activation in different brain areas. However, summarising activation patterns within each area to a single dimension ignores the potential statistical dependencies between their multi-dimensional activity patterns. Representational Connectivity Analyses (RCA) is a method that quantifies the relationship between multi-dimensional patterns of activity without reducing the dimensionality of the data. We consider two variants of RCA. In model-free RCA, the goal is to quantify the shared information for two brain regions. In model-based RCA, one tests whether two regions have shared information about a specific aspect of the stimuli/task, as defined by a model. However, this is a new approach and the potential caveats of model-free and model-based RCA are still understudied. We first explain how model-based RCA detects connectivity through the lens of models, and then present three scenarios where model-based and model-free RCA give discrepant results. These conflicting results complicate the interpretation of functional connectivity. We highlight the challenges in three scenarios: complex intermediate models, common patterns across regions and transformation of representational structure across brain regions. The paper is accompanied by scripts that reproduce the results. In each case, we suggest potential ways to mitigate the difficulties caused by inconsistent results. The results of this study shed light on some understudied aspects of RCA, and allow researchers to use the method more effectively.

show abstract

Perceptual difficulty modulates the direction of information flow in familiar face recognition

Cited by 42 publications

References 82 publications

Reflections and New Perspectives on Face Cognition as a Specific Socio-Cognitive Ability

Reflections and New Perspectives on Face Cognition as a Specific Socio-Cognitive Ability

When the whole is less than the sum of its parts: maximum object category information and behavioral prediction in multiscale activation patterns

Caveats and nuances of model-based and model-free representational connectivity analysis

Contact Info

Product

Resources

About