Modeling face recognition in the predictive coding framework: A combined computational modeling and functional imaging study

A longitudinal multimodal MRI study of the visual network in postoperative delirium

Although structural and functional damage to the brain is considered to be an important neurobiological mechanism of postoperative delirium (POD), alterations in the visual cortical network related to this vulnerability have not yet been determined. In this study, we investigated the impact of alterations in the visual network (VN), as measured by structural and functional magnetic resonance imaging (MRI), on the development of POD. Thirty-six adult patients with frontal glioma who underwent elective craniotomy were recruited. The primary outcome was POD 1-7 days after surgery, as assessed by the Confusion Assessment Method. Cognition before surgery was measured by a battery of neuropsychological tests. Then, we evaluated preoperative and postoperative gray matter volume (GMV) and functional connectivity (FC) alterations by voxel-based morphometry and resting-state functional MRI (rs-fMRI) between the POD and non-POD groups. Multiple logistic regression models were used to investigate the associations between neuroimaging biomarkers and the occurrence of POD. Compared to those in the non-POD group, a decreased GMV in the fusiform gyrus (0.181 [0.018] vs. 0.207 [0.022], FDRp = 0.001) and decreased FC between the fusiform gyrus and VN (0.351 [0.153] vs. 0.610 [0.197], GFRp < 0.001) were observed preoperatively in the POD group, and increased FC between the fusiform gyrus and ventral attentional network (0.538 [0.180] vs. 0.452 [0.184], GFRp =  < 0.001) was observed postoperatively in the POD group. According to our multiple logistic regression analysis, age (Odds ratio [OR]: 1.141 [1.015 to 1.282], P = 0.03) and preoperative fusiform-VN FC (OR 0.001 [0.001 to 0.067], P = 0.01) were significantly related to risk of POD. Our findings suggested that preoperative functional disconnectivity between fusiform and VN might be highly involved in the development of POD. These findings may allow for the discovery of additional underlying mechanisms.

Comprehensive investigation of predictive processing: A cross- and within-cognitive domains fMRI meta-analytic approach

Predictive processing (PP) stands as a predominant theoretical framework in neuroscience. While some efforts have been made to frame PP within a cognitive domain-general network perspective, suggesting the existence of a "prediction network," these studies have primarily focused on specific cognitive domains or functions. The question of whether a domain-general predictive network that encompasses all well-established cognitive domains exists remains unanswered. The present meta-analysis aims to address this gap by testing the hypothesis that PP relies on a large-scale network spanning across cognitive domains, supporting PP as a unified account toward a more integrated approach to neuroscience. The Activation Likelihood Estimation meta-analytic approach was employed, along with Meta-Analytic Connectivity Mapping, conjunction analysis, and behavioral decoding techniques. The analyses focused on prediction incongruency and prediction congruency, two conditions likely reflective of core phenomena of PP. Additionally, the analysis focused on a prediction phenomena-independent dimension, regardless of prediction incongruency and congruency. These analyses were first applied to each cognitive domain considered (cognitive control, attention, motor, language, social cognition). Then, all cognitive domains were collapsed into a single, cross-domain dimension, encompassing a total of 252 experiments. Results pertaining to prediction incongruency rely on a defined network across cognitive domains, while prediction congruency results exhibited less overall activation and slightly more variability across cognitive domains. The converging patterns of activation across prediction phenomena and cognitive domains highlight the role of several brain hubs unfolding within an organized large-scale network (Dynamic Prediction Network), mainly encompassing bilateral insula, frontal gyri, claustrum, parietal lobules, and temporal gyri. Additionally, the crucial role played at a cross-domain, multimodal level by the anterior insula, as evidenced by the conjunction and Meta-Analytic Connectivity Mapping analyses, places it as the major hub of the Dynamic Prediction Network. Results support the hypothesis that PP relies on a domain-general, large-scale network within whose regions PP units are likely to operate, depending on the context and environmental demands. The wide array of regions within the Dynamic Prediction Network seamlessly integrate context- and stimulus-dependent predictive computations, thereby contributing to the adaptive updating of the brain's models of the inner and external world.

Prediction error processing and sharpening of expected information across the face-processing hierarchy

The perception and neural processing of sensory information are strongly influenced by prior expectations. The integration of prior and sensory information can manifest through distinct underlying mechanisms: focusing on unexpected input, denoted as prediction error (PE) processing, or amplifying anticipated information via sharpened representation. In this study, we employed computational modeling using deep neural networks combined with representational similarity analyses of fMRI data to investigate these two processes during face perception. Participants were cued to see face images, some generated by morphing two faces, leading to ambiguity in face identity. We show that expected faces were identified faster and perception of ambiguous faces was shifted towards priors. Multivariate analyses uncovered evidence for PE processing across and beyond the face-processing hierarchy from the occipital face area (OFA), via the fusiform face area, to the anterior temporal lobe, and suggest sharpened representations in the OFA. Our findings support the proposition that the brain represents faces grounded in prior expectations.

Exploring a formal approach to selecting studies for replication: A feasibility study in social neuroscience

Replication of published results is crucial for ensuring the robustness and self-correction of research, yet replications are scarce in many fields. Replicating researchers will therefore often have to decide which of several relevant candidates to target for replication. Formal strategies for efficient study selection have been proposed, but none have been explored for practical feasibility - a prerequisite for validation. Here we move one step closer to efficient replication study selection by exploring the feasibility of a particular selection strategy that estimates replication value as a function of citation impact and sample size (Isager, van 't Veer, & Lakens, 2021). We tested our strategy on a sample of fMRI studies in social neuroscience. We first report our efforts to generate a representative candidate set of replication targets. We then explore the feasibility and reliability of estimating replication value for the targets in our set, resulting in a dataset of 1358 studies ranked on their value of prioritising them for replication. In addition, we carefully examine possible measures, test auxiliary assumptions, and identify boundary conditions of measuring value and uncertainty. We end our report by discussing how future validation studies might be designed. Our study demonstrates the importance of investigating how to implement study selection strategies in practice. Our sample and study design can be extended to explore the feasibility of other formal study selection strategies that have been proposed.