The Relationship between Frontotemporal Effective Connectivity during Picture Naming, Behavior, and Preserved Cortical Tissue in Chronic Aphasia

Abstract

While several studies of task-based effective connectivity of normal language processing exist, little is known about the functional reorganization of language networks in patients with stroke-induced chronic aphasia. During oral picture naming, activation in neurologically intact individuals is found in "classic" language regions involved with retrieval of lexical concepts [e.g., left middle temporal gyrus (LMTG)], word form encoding [e.g., left posterior superior temporal gyrus, (LpSTG)], and controlled retrieval of semantic and phonological information [e.g., left inferior frontal gyrus (LIFG)] as well as domain-general regions within the multiple demands network [e.g., left middle frontal gyrus (LMFG)]. After stroke, lesions to specific parts of the left hemisphere language network force reorganization of this system. While individuals with aphasia have been found to recruit similar regions for language tasks as healthy controls, the relationship between the dynamic functioning of the language network and individual differences in underlying neural structure and behavioral performance is still unknown. Therefore, in the present study, we used dynamic causal modeling (DCM) to investigate differences between individuals with aphasia and healthy controls in terms of task-induced regional interactions between three regions (i.e., LIFG, LMFG, and LMTG) vital for picture naming. The DCM model space was organized according to exogenous input to these regions and partitioned into separate families. At the model level, random effects family wise Bayesian Model Selection revealed that models with driving input to LIFG best fit the control data whereas models with driving input to LMFG best fit the patient data. At the parameter level, a significant between-group difference in the connection strength from LMTG to LIFG was seen. Within the patient group, several significant relationships between network connectivity parameters, spared cortical tissue, and behavior were observed. Overall, this study provides some preliminary findings regarding how neural networks for language reorganize for individuals with aphasia and how brain connectivity relates to underlying structural integrity and task performance.

Keywords: aphasia; behavioral performance; cortical damage; dynamic causal modeling; effective connectivity; fMRI; oral picture naming.

FIGURE 1

Schematic of the fMRI picture naming task.

FIGURE 2

Lesion overlap of all thirteen PWA included in the DCM analysis.

FIGURE 3

Overview of the sequence of (A) fMRI and (B) DCM methods.

FIGURE 4

DCM model space. Full, bidirectional endogenous connections between all regions were modeled in DCM-A. For each model, driving input to only one region was modeled in DCM-C. All possible combinations of uni- and bidirectional modulations were modeled across the model space; for each model, the input region modulated at least one other region in DCM-B. The full model space for all 24 models in Family #1 is schematized in the figure above (1). Family #2 included models with the same modulatory connections as Family #1 with three additional models (2) and excluding models #1, #4, and #7 due to lack of modulation from LMFG to the other two regions. Similarly, Family #3 included models with the same modulatory connections as Family 1 with three additional models (3) and excluding models #9, #10, and #11 due to lack of modulation from LMTG to the other two regions.

FIGURE 5

Whole brain activation. (A) Results of the one-sample t-test in PWA at uncorrected (t = 3.05, p < 0.005) for pictures > scrambled pictures. (B) Results of the one-sample t-test in controls at uncorrected (t = 3.25, p < 0.005) for pictures > scrambled pictures. (C) Overlap of the 13 individual PWA activation maps at uncorrected (p < 0.001), cluster size of 3 voxels for pictures – scrambled pictures. (D) Overlap of the 10 individual control activation maps at uncorrected (p < 0.001), cluster size of 5 voxels pictures – scrambled pictures.

FIGURE 6

Family wise BMS. (A) Group-level family wise BMS results. (B) Single-subject family wise BMS for the PWA.

FIGURE 7

Correlations between percentage of spared tissue and strength of the connections (i.e., Ep.B values in Hz). (A) For family #1, significant correlations were found between the connection strength of LMFG → LIFG and the percentage of spared tissue in LIFG (shown on the left) and LMFG (shown on the right). (B) For family #2, a significant correlation was found between the connection strength of LMTG → LIFG and the amount of spared tissue in LMTG. (C) For family #3, significant correlations were found between the connection strength of LMFG → LMTG and percentage spared tissue in LMFG as well as the connection LMTG → LMFG and the amount of spared tissue in LMTG.

FIGURE 8

Correlations between percentage of spared tissue and strength of task-induced perturbation to specific regions (i.e., Ep.C values in Hz). (A) For family #1, an association that approached significance was found between strength of driving input of LIFG and amount of spared tissue in LIFG. (B) For family #3, a trending association between driving input strength of LMTG and the amount of spared tissue in LMTG.

FIGURE 9

Correlations between behavioral performance and strength of the connections (i.e., Ep.B values in Hz). (A) For family #1, significant correlations were found between the connection strength of LMFG → LMTG and behavioral accuracy on the naming screener (shown on the left) and the fMRI task (shown on the right). (B) For family #2, a significant correlation was found between the connection strength of LIFG → LMTG and fMRI task accuracy. (C) For family #3, a significant correlation was found between the connection strength of LMFG → LMTG and the average naming screener accuracy.

FIGURE 10

Correlations between behavioral accuracy and strength of task-induced perturbation to specific regions (i.e., Ep.C values in Hz). (A) For family #1, significant associations were found between strength of task-induced perturbation to LIFG and accuracy on the naming screener (shown on the left) and on the fMRI task (shown on the right). (B) For family #2, a significant relationship was seen between strength of driving input to LMFG and fMRI task accuracy.