In a groundbreaking exploration of how human brains integrate the senses of vision and touch, researchers have uncovered new evidence highlighting the intricate somatotopic organization within the dorsolateral visual cortex. This discovery sheds light on the fundamental mechanisms through which visual and somatosensory systems communicate and coordinate to create a seamless perception of our own bodies and the surrounding environment. The study probes how the brain’s visual areas, conventionally understood through retinotopy—the mapping of visual input from the retina—can also embody somatotopic maps that reflect the bodily layout, revealing a remarkable neural alignment between seeing and feeling.
The fundamental question driving this research centers on whether and how visual and tactile systems share common spatial reference frames in the cortex. Notably, the visual system’s well-established retinotopic organization presents an elegant coordinate frame mapped from the retina’s surface onto the brain. Could the somatotopic map, detailing the body’s sensory surface, be aligned with retinotopic maps to facilitate ecological interactions such as the alignment of vision with bodily parts positioned in specific visual fields? For instance, feet generally occupy the lower visual field, suggesting a natural concordance in sensory mapping that could optimize processing for environmental affordances, such as foot placement or navigating obstacles.
To interrogate this possibility, the research team conducted a detailed permutation-based searchlight analysis targeting correlations between somatotopic organization and vertical visual field preferences derived from the Human Connectome Project’s comprehensive retinotopy dataset. The analysis revealed compelling evidence for this alignment primarily within dorsal visual regions, including retinotopically defined area V3B and lateral occipital regions such as LO1 and the superior extrastriate body area (EBA). These areas are known to participate robustly in the dorsal visuomotor stream, which governs spatial processing and motor planning, affirming the hypothesis that linked somatotopic-retinotopic frames underpin sensorimotor integration in these dorsal zones.
Intriguingly, this alignment was absent in more ventral regions encompassing the fusiform body area (FBA), inferior EBA, and the visual word form area (VWFA). These regions exhibit a pronounced foveal bias, reflecting their specialization in high-acuity central vision and object recognition rather than broad visuospatial mapping. Yet, despite the absence of retinotopic-somatotopic alignment, the VWFA and adjacent areas demonstrate significant links to non-textual gestural processing, particularly for hands and faces, indicating a different mode of multisensory convergence that transcends pure spatial correspondence.
This observation led the researchers to propose a second hypothesis: somatotopic organization in ventral lateral visual cortex corresponds to a categorical level of visual body-part selectivity rather than positional retinotopic alignment. To test this, the team harnessed the power of the Natural Scenes Dataset (NSD), an extensive 7 Tesla fMRI dataset containing participant responses to thousands of natural images, many featuring diverse human postures and body parts. Utilizing advanced pose-estimation algorithms, researchers extracted 17 key anatomical body points from these images to construct a forward model predicting single-voxel visual selectivity to individual body parts.
The modeling illuminated that somatotopic maps in ventral regions, spanning from superior FBA through ventral EBA and into the VWFA, indeed predict visual selectivity for specific body parts. This reveals a semantic-level organization mirroring bodily anatomy that manifests in visual cortex areas engaged in detailed body recognition and categorical processing. The ventral visual stream, therefore, embodies a form of “visual body-part tuning” that reflects consistent coding with somatosensory maps, aligning the perception of body parts in vision with their tactile representations.
Collectively, these findings delineate two complementary forms of multisensory alignment in the human brain’s visual cortex. The dorsal stream exhibits visuospatial alignment linking retinotopic and somatotopic frames, thereby supporting sensorimotor computations closely tied to spatial localization and action planning. Meanwhile, the ventral stream engages semantic alignment, integrating categorical visual body-part selectivity anchored in somatotopic organization, thereby supporting identity recognition and perceptual categorization of human form.
This dual alignment underscores the pivotal role of the lateral visual cortex as a neural interface bridging dorsal “where/how” and ventral “what” pathways. The lateral visual cortex’s capacity to integrate both visuospatial and semantic dimensions enables it to support rich, contextually nuanced body representations that facilitate complex behaviors such as gesture recognition, object manipulation, and social communication.
Methodologically, the study leveraged state-of-the-art neuroimaging techniques including high-resolution functional MRI data sets enriched by naturalistic stimuli and computational modeling. The use of permutation-based searchlight analyses allowed refined, spatially precise detection of map congruences across cortical surfaces. This approach exemplifies the power of combining neuroimaging with machine learning and computer vision to decode the brain’s multisensory representations with unprecedented granularity.
From a broader perspective, these insights elevate our understanding of how the brain constructs unified body maps by synthesizing multiple sensory modalities. Such knowledge bears significant relevance for the design of neuroprosthetics, brain-machine interfaces, and rehabilitation strategies aimed at restoring sensorimotor functions after injury or disease. Mapping the neural substrates of multisensory integration opens paths toward technologies that more effectively emulate the brain’s natural coordination of vision and touch.
Moreover, the demonstration that visual areas extrapolate bodily information based on contextual cues found in natural scenes enhances current models of embodied cognition, emphasizing that body representation in the brain is not simply tactile but also visually informed and contextually plastic. This has ramifications for research into social cognition, body ownership illusions, and disorders of body representation such as phantom limb syndrome.
Looking forward, future research may delve deeper into the temporal dynamics underpinning the interplay between somatosensory and visual representations, employing techniques such as magnetoencephalography (MEG) or intracranial recordings. Investigations at the cellular and circuit levels could clarify the synaptic and network mechanisms enabling these cross-modal alignments. Furthermore, expanding such research across developmental stages or clinical populations could illuminate how these multisensory maps emerge and adapt throughout life.
In sum, this novel research unveils a sophisticated architecture of vicarious body maps that bridge vision and touch in the human brain. The dual reference frames identified articulate how spatial and semantic dimensions converge in the visual cortex, offering a comprehensive framework for understanding the neural basis of embodied perception and multisensory integration. This represents a major advance in cognitive neuroscience, charting new avenues for interdisciplinary exploration of body representation.
The study’s implications resonate deeply within neuroscience and beyond, revealing the brain’s extraordinary capacity to fuse different senses into coherent, actionable body maps. Such insights not only deepen our grasp of fundamental brain functions but also inspire innovative approaches to medical technology, artificial intelligence, and human-computer interaction that emulate the brain’s seamless integration of sight and touch.
Subject of Research: Integration of visual and somatosensory maps in the human brain’s visual cortex
Article Title: Vicarious body maps bridge vision and touch in the human brain
Article References:
Hedger, N., Naselaris, T., Kay, K. et al. Vicarious body maps bridge vision and touch in the human brain. Nature (2025). https://doi.org/10.1038/s41586-025-09796-0
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41586-025-09796-0
Keywords: somatotopic organization, retinotopy, dorsal visual cortex, ventral visual cortex, multisensory integration, body representation, fMRI, natural scenes dataset, pose estimation, neural mapping
