Frontoparietal contribution for conscious visual representations and related perceptual confidence

2470 

Abstract Submission 

J. Amaro-Fuenzalida^1,2, Tomás Ossandón Valdés^3,4, Aurelio Cortese⁵, Ranganatha Sitaram⁶

¹Centro Interdisciplinario de Neurociencia, Pontificia Universidad Católica de Chile, Medicina, Santiago de Chile, Chile, ²Neurodynamic of Cognition Laboratory, Departamento de Psiquiatría, Pontificia Universidad Católica de Chile, Santiago, Chile, ³Neurodynamic of Cognition Laboratory, Departamento de Psiquiatría, Pontificia Universidad Católica d, Santiago de Chile, Chile, ⁴Centro Interdisciplinario de Neurociencia, Pontificia Universidad Católica de Chile, Medicina, Santiago, Chile, ⁵Department of Decoded Neurofeedback, ATR Computational Neuroscience Laboratories, Kyoto, Japan, ⁶St. Jude Children's Research Hospital, Memphis, TN

J. Amaro-Fuenzalida, PhD(c)  
Centro Interdisciplinario de Neurociencia, Pontificia Universidad Católica de Chile, Medicina|Neurodynamic of Cognition Laboratory, Departamento de Psiquiatría, Pontificia Universidad Católica de Chile
Santiago de Chile, Chile|Santiago, Chile

Tomás Ossandón Valdés, PhD  
Neurodynamic of Cognition Laboratory, Departamento de Psiquiatría, Pontificia Universidad Católica d|Centro Interdisciplinario de Neurociencia, Pontificia Universidad Católica de Chile, Medicina
Santiago de Chile, Chile|Santiago, Chile

Aurelio Cortese, PhD  
Department of Decoded Neurofeedback, ATR Computational Neuroscience Laboratories
Kyoto, Japan

Ranganatha Sitaram  
St. Jude Children's Research Hospital
Memphis, TN

Consciousness as content refers to the experience of objects, commonly known as awareness (Searle, 2007). The Global Neuronal Workspace Model (GNWM) suggests that perceptual awareness relies on the activation of the fronto-parietal network (Mashour, 2020). Evidence indicates a frontal contribution to the awareness process when subjects report their experiences (Del Cul, 2007). However, GNWM does not explain how conscious contents are represented but focuses on the mechanisms required to access them. Theories like High Order Theory (HOT) address representation, proposing frontal meta-representation of posterior activity (Mashour, 2020).

fMRI and neurofeedback experiments suggest perceptual content representations associate with posterior regions (Amano, 2017; Shibata, 2011). It remains unclear whether the representation of perceptual content is directly linked to frontal regions per se, even when reporting is necessary. Multi-Voxel Pattern Analysis techniques have identified that perceptual confidence can be decoded in frontal regions (Cortese, 2017) but objective properties of stimuli seem linked to posterior regions. This research aims to determine the frontoparietal contribution to conscious visual representations and its connection with perceptual confidence.

fMRI data from 26 participants performing the Shadlen Random Dot Motion task (Cortese, 2017) were collected. We presented a set of random dots on a screen with a percentage of dots moving in a specific direction (coherence). We asked our participants to identify movement direction (objective rating) and perceptual confidence (subjective rating). Each participant underwent at least 160 trials. Each stimulus was calibrated under coherence threshold estimated for each participant by Quest Psychtoolbox function. Trials comprised Direction, Coherence, Score, and Confidence. Two fMRI analyses were conducted: I. ROI analysis, subtracting estimated beta values for dichotomic conditions (e.g. below or above threshold). II. Multi-Voxel Pattern Analysis, decoding brain activity with a Searchlight approach through The Decoding Toolbox, comparing voxel accuracy between lobes using the Desikan-Killany-Tourville atlas (Hatamimajoumerd, 2022).

Data collection and sharing for this project was provided by the DecNef Department at the Advanced Telecommunication Research Institute International, Kyoto, Japan. FONDECYT regular 1230383

ROI analysis revealed significant differences in beta ROI activity between stimuli with 25% and 75% coherence as a function of topography (p <0.05). The difference between frontal and posterior activity under low and high coherence was significantly higher. The left hemisphere also showed significant differences. Right hemisphere differences were observed between ROIs for conscious and nonconscious conditions (p=0.0257, p=0.012). The most accurate ROI for decoding movement direction below the awareness threshold included frontal regions. Significant differences in right hemisphere accuracy were observed in unconscious direction trials (p=0.023), while the left hemisphere showed differences between ROI accuracy for Score and Confidence.

We expected to find high contribution when participants were aware of stimuli properties, while results suggest frontal region contribution, especially under low signal strength (low coherence). Decoding movement direction below the consciousness threshold highlighted the ventro-medial prefrontal cortex (vmPFC) significant contribution. This suggests frontal activity in perceptual tasks relates to working memory, acting as a memory buffer under conditions requiring more time for representation. This has been observed through no-report paradigm (Hatamimajoumerd, 2022). It is also possible that conscious experience about complex stimuli may be internally driven by priors (LeDoux & Lau, 2020) or imagination when stimuli is not strong enough to surpass subjective thresholds (Dijkstra & Fleming, 2023).

Higher Cognitive Functions Other ²

Consciousness and Awareness ¹

Consciousness

FUNCTIONAL MRI

Perception