Back

The hierarchical organization of human praxis: A lesion-connectome mapping study of limb and oral apraxias

Nishio, Y.; Itabashi, R.; Kataoka, Y.; Yazawa, Y.; Mori, E.

2025-12-14 neuroscience
10.64898/2025.12.13.694084 bioRxiv
Show abstract

Background and ObjectivesTo characterize the anatomical network architectures underlying effector-specific apraxias, specifically comparing the neuroanatomical substrates of limb apraxia (LA), buccofacial apraxia (BFA), and apraxia of speech (AOS). MethodsWe evaluated 136 patients with acute left-hemisphere ischemic stroke for the presence of LA, BFA, and AOS. We used voxel-based, region-of-interest-based, and connectome-based lesion-symptom mapping to identify the cortical lesions and white matter disconnections associated with each apraxia subtype. ResultsLA was associated with damage to the inferior parietal lobule and widespread white matter disconnections, including intrahemispheric temporo-parietal pathways and interhemispheric transcallosal fibers. In contrast, BFA was associated with localized anterior lesions centered on the mid-lower portion of the left precentral gyrus and adjacent frontal operculum, without massive long-range white matter disconnections. While primary analyses for AOS yielded no suprathreshold clusters, likely washed out by the spatial variance of lesions causing concomitant aphasia, exploratory analyses localized AOS to the mid-portion of the precentral gyrus, situated slightly dorsal to the core region of BFA. DiscussionLA and oral apraxias differ fundamentally in their underlying network architectures rather than merely their effectors. LA involves the disruption of a distributed, bilaterally integrated temporo-parietal network. Conversely, oral apraxias depend on localized anterior networks, with a dorso-ventral dissociation within the precentral gyrus distinguishing pure AOS (dorsal primary motor) from BFA (ventral premotor). These findings highlight how the brain uses distinct network principles, distributed versus localized, to control different classes of skilled action.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.