Delineating Functional Subregions of the Left Precentral Gyrus: A fMRI Meta-Analysis

2109 

Abstract Submission 

Yiming Zhang¹, Chu-Chung Huang¹

¹East China Normal University, Shanghai, China

Yiming Zhang  
East China Normal University
Shanghai, China

Chu-Chung Huang  
East China Normal University
Shanghai, China

The human precentral gyrus (PCG), known as the primary motor cortex, is crucial for voluntary movement control. It connects to the spinal cord via pathways like the corticospinal and corticobulbar tracts, with damage leading to motor deficits. Recent studies, however, reveal its roles in speech production (Silva et al., 2022) and motor learning (Rubin et al., 2022), challenging its traditional view as solely a motor center. This study revisits the PCG from the perspective of involved cognitions. By employing a meta-analytic approach at PCG region, we aim to parcellate PCG into subregions with distinct cognitive functions. Furthermore, HCP-released diffusion MRI data were used to examine the structural connectivity pattern among the identified subregions. We hypothesize that 1) PCG can be separated into anterior and posterior bank; 2) PCG may have a distinct dorsal-ventral parcellation in the anterior segment.

We searched for task-based fMRI studies on the Precentral Gyrus using Brainmap, focusing on the left hemisphere and studies from January 2010 to July 2022. Criteria excluded subjects outside 18-65 years or with illnesses, yielding 162 studies. Studies were categorized by functions as 'emotion', 'language', 'social cognition', 'execution', 'attention', 'working memory', and 'movement', each with at least 15 experiments. To complement ROI-based analysis, we conducted a whole-brain meta-analysis (Gentili et al., 2019), selecting studies reflecting specific behaviors. Following the previous steps, we selected 7 groups of studies. Using Brainmap GingerALE, we performed an ALE meta-analysis (Turkeltaub et al., 2002), transforming foci into a probability distribution for each voxel. Settings included FWE correction 0.01 and 10,000 permutations. ALE maps were registered to cvs_avg35_inMNI152, providing ALE values for each vertex in the left hemisphere. We clustered vertices in the left PCG based on ALE values, including 'parsopercularis' for its language function relevance. Clustering used k-means with cityblock distance, evaluated by SSE, silhouette coefficient, and Calinski-Harabasz index. To assess structural bases, we conducted diffusion tractography on HCP 7T dataset DWI data. Steps included converting the aparc+aseg atlas, transforming it to subject-specific space, generating a connectivity matrix, and comparing connectivity patterns between clusters for all subjects.

In our parcellation and function decoding analysis, we found that a four-cluster solution was optimal for the left PCG. This led to identifying four distinct segments: PP, VP, DP, and the other region. MACM revealed distinct coactivation patterns for these clusters. PP showed increased coactivation in sensorimotor areas, VP in language processing and action observation regions, and DP in eye field and execution-related areas. Diffusion tractography analysis highlighted significant connectivity differences among these clusters. DP showed stronger connectivity to the left caudal middle frontal region compared to PP, while VP had enhanced connectivity to the left pars opercularis and supramarginal regions. PP, in contrast, demonstrated stronger links with sensorimotor areas like the left postcentral, putamen, and pallidum.

Our meta-analysis offers detailed insights into the left PCG, identifying 3 unique subregions: PP, similar to M1 in function and structure, and the anterior VP and DP, each with distinct roles. PP is engaged in basic cognitive tasks, while VP focuses on language, and DP is associated with action observation, highlighting their diverse cognitive functions and networks.

Cortical Anatomy and Brain Mapping ¹

Brain Atlases ²

Emotions

Language

Meta- Analysis

Motor

Social Interactions