Development of information sharing in language neocortex in childhood‐onset drug‐resistant epilepsy

Abstract

Summary Objective We studied age‐related dynamics of information sharing among cortical language regions with electrocorticographic high‐gamma modulation during picture‐naming and story‐listening tasks. Methods Seventeen epilepsy patients aged 4‐19 years, undergoing extraoperative monitoring with left‐hemispheric subdural electrodes, were included. Mutual information (MI), a nondirectional measure of shared information, between 16 pairs of cortical regions of interest, was computed from trial‐averaged 70–150 Hz power modulations during language tasks. Impact of age on pairwise MI between language regions and their determinants were ascertained with regression analysis. Results During picture naming, significant increase in MI with age was seen between pairwise combinations of Broca's area, inferior precentral gyrus (iPreC), and frontal association cortex (FAC); Wernicke's area and posterior association cortex (PAC); and Broca's and Wernicke's areas. During story listening, significant age‐related increase in MI was seen between Wernicke's area and either Broca's area, FAC, or PAC; and between Broca's area and FAC. Significant impact of baseline intelligence quotient was seen on the relationship between age and MI for all pairs, except between Broca's area and iPreC. The mean MI was higher during naming compared to listening for pairs including iPreC with Broca's area, FAC, or PAC and was lower for pairs of Wernicke's area or PAC with anterior language regions. Significance Information sharing matures with age “within” frontal and temporoparietal language cortices, and “between” Broca's and Wernicke's areas. This study provides evidence for distinct patterns of developmental plasticity within perisylvian language cortex and has implications for planning epilepsy surgery.

Neuronal Circuits Supporting Development of Visual Naming Revealed by Intracranial Coherence Modulations

“…By having the same starting set of images across all patients, eliminating pictures that the patient was unable to name, and recording under similar environment, we have tried to mitigate the effects of word usage, accuracy, and stimulus context. Response latency for picture naming is also influenced by literacy levels, and mutual information in HGM between parcel pairs is affected by IQ ( Reis et al, 2001 ; Arya et al, 2019c ). Therefore, we excluded patients with VCI < 70.…”

Deep reasoning neural network analysis to predict language deficits from psychometry‐driven DWI connectome of young children with persistent language concerns

“…For the differences in edges that are predictive of expressive vs. receptive scores, our Grad‐CAM analysis found that there were commonalities between edges associated with expressive and receptive scores: they both involved bilateral structures and had one common structure (left angular gyrus). Significant information sharing was also identified by language‐associated electrocorticography high‐gamma modulation “within” frontal and temporoparietal language cortices, and “between” classical language areas in the dominant hemisphere (i.e., Broca's and Wernicke's areas) (Arya et al, 2019), suggesting that expressive and receptive language functions may communicate and share neural activity across multiple language regions. Interestingly, our result provided neurobiological evidence on a spatial redundancy (Schomers, Garagnani, & Pulvermüller, 2017) determined by our dilated CNN + RN to be most predictive of expressive and receptive scores (e.g., left ANG).…”

6D dynamic tractography visualizes the effective connectivity supporting specific linguistic stages in the developing brain

“…While task-based functional connectivity, CCSR-based effective connectivity, and DWI-based structural connectivity have speci c strengths, each method has an inevitable limitation. Previous studies using Granger causality reported that cortical sites showing task-related highgamma augmentation during a shared time window could propagate neural activity between them 12,13,20,114 . However, Granger causality cannot consistently distinguish direct versus indirect connectivity.…”

“…Our 6D dynamic tractography approach integrates all three modalities into a single model and visualization, which revealed the white matter pathways supporting rapid neural propagations between sites engaged to speci c linguistic stages. We hypothesize that the dynamic trajectories observable via 6D tractography are more likely to be used for auditory naming if the remote areas connected are simultaneously engaged in the same linguistic stage 12,13,20,114 . Indeed, we found that early CCSRs were larger when the stimulus and recording sites showed high-gamma augmentation during the same linguistic stage.…”