Publicação
Unravelling the neural based of material and object perception
| Resumo: | The human visual system allows us to rapidly recognize both objects and the materials they are made of—a crucial ability for interacting with our environment. While object recognition has been extensively studied, the neural mechanisms underlying material perception—such as how we process texture, gloss, and rigidity—and material recognition—that something is made of wood and not plastic—are less understood. This thesis investigates how the brain distinguishes between object form and material composition, combining neuroimaging and neuropsychological evidence. We conducted a whole-brain univariate functional magnetic resonance imaging (fMRI) study with twenty-two healthy adults, using carefully controlled visual stimuli to independently probe material and object-shape processing. Our analyses revealed that medial ventral stream regions—including the collateral sulcus (CoS) and parahippocampal gyrus (PHG)—selectively process material properties, whereas lateral occipitotemporal regions—such as the lateral occipital complex (LOC)—are more involved in object-shape recognition. Using whole-brain mass-univariate contrasts, we establish a clear double dissociation: material information concentrates in ventromedial occipitotemporal cortex (lingual gyrus and collateral sulcus), whereas shape robustly engages LOC alongside dorsal visuospatial regions (IPS/SPL, V7). Critically, directly comparing materials with matched texture patches revealed no region preferring Texture > Material at corrected thresholds; instead, Material > Texture highlighted anterior collateral sulcus and parahippocampal gyrus (plus left V7/V3A and left lingual gyrus), consistent with a posterior-to-anterior progression from shared surface/texture encoding to more context-enriched material representations. Category-specific overlays clarify how material coding interfaces with object systems: for manipulable objects (tools), material-selective cortex in anterior/medial CoS spatially overlaps the tool network, whereas the Material > Tools advantage shifts posteriorly (lingual/posterior CoS) and into left PHG—suggesting that tool images already strongly engage anterior CoS. In contrast, place processing (PPA) lies largely anterior and non-overlapping with the material pathway, with any convergence restricted to posterior CoS. Together, these findings support partly segregated pathways for surface/material properties and geometric form, a hierarchical material stream culminating in object-linked codes, and testable neuropsychological predictions—for example, that damage centered on anterior collateral sulcus would disproportionately impair tool-related material judgments while sparing scene recognition. |
|---|---|
| Autores principais: | Ramdani, Oualid |
| Assunto: | Processamento de Forma Análise de fMRI Reconhecimento de Objetos fMRI Analysis Object Recognition Material Perception Ventral Visual Pathway Shape Processing Percepção de Materiais Via Visual Ventral |
| Ano: | 2025 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Coimbra |
| Idioma: | inglês |
| Origem: | Estudo Geral - Universidade de Coimbra |
| Resumo: | The human visual system allows us to rapidly recognize both objects and the materials they are made of—a crucial ability for interacting with our environment. While object recognition has been extensively studied, the neural mechanisms underlying material perception—such as how we process texture, gloss, and rigidity—and material recognition—that something is made of wood and not plastic—are less understood. This thesis investigates how the brain distinguishes between object form and material composition, combining neuroimaging and neuropsychological evidence. We conducted a whole-brain univariate functional magnetic resonance imaging (fMRI) study with twenty-two healthy adults, using carefully controlled visual stimuli to independently probe material and object-shape processing. Our analyses revealed that medial ventral stream regions—including the collateral sulcus (CoS) and parahippocampal gyrus (PHG)—selectively process material properties, whereas lateral occipitotemporal regions—such as the lateral occipital complex (LOC)—are more involved in object-shape recognition. Using whole-brain mass-univariate contrasts, we establish a clear double dissociation: material information concentrates in ventromedial occipitotemporal cortex (lingual gyrus and collateral sulcus), whereas shape robustly engages LOC alongside dorsal visuospatial regions (IPS/SPL, V7). Critically, directly comparing materials with matched texture patches revealed no region preferring Texture > Material at corrected thresholds; instead, Material > Texture highlighted anterior collateral sulcus and parahippocampal gyrus (plus left V7/V3A and left lingual gyrus), consistent with a posterior-to-anterior progression from shared surface/texture encoding to more context-enriched material representations. Category-specific overlays clarify how material coding interfaces with object systems: for manipulable objects (tools), material-selective cortex in anterior/medial CoS spatially overlaps the tool network, whereas the Material > Tools advantage shifts posteriorly (lingual/posterior CoS) and into left PHG—suggesting that tool images already strongly engage anterior CoS. In contrast, place processing (PPA) lies largely anterior and non-overlapping with the material pathway, with any convergence restricted to posterior CoS. Together, these findings support partly segregated pathways for surface/material properties and geometric form, a hierarchical material stream culminating in object-linked codes, and testable neuropsychological predictions—for example, that damage centered on anterior collateral sulcus would disproportionately impair tool-related material judgments while sparing scene recognition. |
|---|
Atividades financiadas
Não foram encontrados projetos financiados.