The key finding
A 2026 meta-analysis of 46 brain imaging studies involving 591 participants has revealed that mental imagery is not a single brain process, but rather splits into two distinct neural pathways depending on what you’re imagining. When people visualize static objects in detail—like picturing the texture of a brick wall—activity concentrates in left-lateralized frontoparietal regions and the ventral visual stream. In contrast, when people imagine spatial transformations—such as mentally rotating furniture or navigating a route—bilateral superior parietal lobules and the dorsal visual stream light up. This finding suggests that mental imagery mirrors the same “what” versus “where/how” pathways your brain uses during actual perception.
What the study looked like
Researchers conducted a pre-registered activation likelihood estimation (ALE) meta-analysis, a statistical method that pools results from multiple brain imaging studies to identify consistent patterns of activation. They analyzed 46 fMRI and PET studies published through 2026, encompassing 591 total participants and 772 activation foci across the brain. The studies examined two types of mental imagery tasks: object imagery (visualizing static objects with rich visual detail) and spatial imagery (visualizing dynamic transformations, such as mental rotation or spatial navigation). Conjunction analyses identified brain regions active during both types of imagery, while contrast analyses pinpointed regions uniquely associated with each imagery type. This approach allowed the team to map out overlapping versus distinct neural signatures for different forms of mental visualization.
Why researchers think this happened
The authors propose that mental imagery recruits the same dual-stream architecture that supports visual perception. In vision science, the ventral stream (running from occipital to temporal cortex) processes object identity—answering “what is it?”—while the dorsal stream (extending to parietal regions) handles spatial location and action guidance—answering “where is it?” and “how do I interact with it?” The current findings suggest this division extends to internally generated images. Object imagery, which emphasizes detailed visual features like color and shape, appears to engage ventral “what” pathway structures. Spatial imagery, which involves mentally manipulating objects in three-dimensional space, recruits dorsal “where/how” regions linked to spatial transformations and motor planning. Both types share a common frontoparietal network (medial frontal gyrus, precuneus, inferior parietal lobule), likely reflecting general attentional and working memory demands inherent to any imagery task.
How to read this carefully
This is a meta-analysis, meaning it synthesizes existing studies rather than collecting new data—results depend on the quality and heterogeneity of included papers. The 46 studies varied in task design, scanner parameters, and participant demographics, which may introduce noise. Because meta-analyses identify average activation patterns across studies, individual differences in imagery ability or strategy may be obscured. Importantly, this work shows neural associations, not causal mechanisms; we cannot yet say whether activating these pathways directly produces imagery or whether they are downstream effects. The sample size of 591 is substantial for neuroimaging, but replication in larger, more diverse cohorts would strengthen confidence. Finally, the study does not address how these pathways develop or how imagery ability varies across age, culture, or expertise.
What this means for everyday life
Given these findings, it might be worth considering how you use mental imagery in daily tasks. If your work involves design, architecture, or mechanical troubleshooting—activities requiring you to mentally rotate or manipulate objects—you may be relying heavily on dorsal stream spatial imagery. Conversely, if you’re an artist visualizing color palettes or a writer conjuring detailed character appearances, you might engage ventral stream object imagery more intensely. Educators and trainers could explore whether matching instructional methods to imagery type (e.g., spatial diagrams for navigation tasks, detailed illustrations for object recognition) improves learning outcomes. For individuals exploring cognitive strengths, recognizing whether you naturally excel at detailed visualization versus spatial transformation could inform career choices or skill development strategies. This research also invites curiosity about how practices like meditation, which often involve sustained imagery, might differentially engage these two systems.