Neuroplasticity in Aphasia: A Proposed Framework of Language Recovery
Publication: Journal of Speech, Language, and Hearing Research
Volume 62, Number 11
Pages 3973-3985
Abstract
Purpose
Despite a tremendous amount of research in this topic, the precise neural mechanisms underlying language recovery remain unclear. Much of the evidence suggests that activation of remaining left-hemisphere tissue, including perilesional areas, is linked to the best treatment outcomes, yet recruitment of the right hemisphere for various language tasks has also been linked to favorable behavioral outcomes. In this review article, we propose a framework of language recovery that incorporates a network-based view of the brain regions involved in recovery.
Method
We review evidence from the extant literature and work from our own laboratory to identify findings consistent with our proposed framework and identify gaps in our current knowledge.
Results
Expanding on Heiss and Thiel's (2006) hierarchy of language recovery, we identify 4 emerging themes: (a) Several bilateral regions constitute a network engaged in language recovery; (b) spared left-hemisphere regions are important components of the network engaged in language recovery; (c) as damage increases in the left hemisphere, activation expands to the right hemisphere and domain-general regions; and (d) patients with efficient, control-like network topology show greater improvement than patients with abnormal topology. We propose a mechanistic model of language recovery that accounts for individual differences in behavior, network topology, and treatment responsiveness.
Conclusion
Continued work in this topic will lead us to a better understanding of the mechanisms underlying language recovery, biomarkers that influence recovery, and, consequently, more personalized treatment options for individual patients.
Presentation Video
Understanding the nature of language recovery in aphasia is one of the fundamental and momentous questions in aphasia research as it impacts the diagnosis, treatment, and prognosis of patients with aphasia (PWA). A significant body of research has enumerated the nature of language recovery in the acute, subacute, and chronic phases of aphasia; for instance, Kiran and Thompson (2019) have reviewed the underlying neuroplastic mechanisms supporting treatment-related changes in the brain. In this review article, we refrain from repeating a review of the neuroimaging in aphasia literature already done by Kiran and Thompson (2019). Instead, this review article is focused on four emerging observations in studies examining language recovery in aphasia. We end this review article with a proposed mechanistic model of language recovery that accounts for individual differences in behavior, network topology, and treatment responsiveness.
The schematic in Figure 1 integrates work from several key articles (Hillis & Heidler, 2002; Saur et al., 2006) to summarize the presumed trajectory of language recovery after stroke. In the hours and days after a stroke, the degree of tissue reperfusion and reinstatement of functioning of the ischemic penumbra is a critical factor in determining the degree of recovery. In the weeks and months after the stroke, reorganization of brain structure–function relationships appears to determine the degree of recovery that is likely to occur. As recovery progresses into the chronic phase, months and years after the stroke, it seems that establishing new pathways and developing compensatory mechanisms for lost function ultimately determine the degree of long-term recovery. Notably, Saur et al. (2006) found that reduced activation of remaining left-hemisphere (LH) areas in the acute phase was followed by an upregulation of recruitment of right hemisphere (RH) homologues, which was, in turn, associated with language recovery in subacute aphasia. While several patients recovered their language skills by 1 year in this study, improvement in language function was associated with an upregulation and normalization of LH activation.
Thus, while it is well understood that the brain undergoes tremendous spontaneous recovery in the first few months after the stroke (see Kiran & Thompson, 2019), language recovery continues even after the initial 6–9 months postonset and well into the chronic stage. What this emerging evidence indicates is that recovery in the subacute and chronic phases involves a complex interplay between residual LH and intact RH regions that involves reorganization of connections in order to promote improved language recovery over time. Thus far, studies provide strong evidence for the above claim, but a structured mechanism as to how this occurs has remained elusive. In this review article, we propose a mechanistic model of language recovery expanding from a similar hierarchy proposed by Heiss and Thiel (2006) but that focuses on individual differences in recovery. In their framework, Heiss and Thiel suggested that complete or optimal language recovery occurred when there was minor damage to LH regions not central to language processing. Second, satisfactory (but incomplete) language recovery would be observed when there was damage to core LH language regions but residual ipsilesional tissue remained functional, resulting in intrahemispheric compensation. Finally, poor or limited language would be observed when there was extensive damage to the entire LH, leaving only homotopic contralesional RH regions available for language recovery.
Using the Heiss and Thiel (2006) model as a starting point, we identify four emerging themes in the literature examining language recovery that ultimately postulates that the degree to which this network topology instantiates in individual patients determines the extent of recovery for that patient. While Heiss and Thiel posit a hierarchical recovery, we suggest that bilateral regions may be dynamically recruited based on the size and location of individual lesions. Furthermore, to the extent that LH regions are spared, they are important and probably dominant parts of the network engaged in language recovery. Another emerging finding is that, as damage increases in the LH, activation expands to the RH and multiple-demand (MD) or domain-general regions not traditionally involved in language. Finally, emerging evidence suggests that the consequent network reorganization may determine the extent of language recovery patients ultimately show. Overall, patients with network topology closer to a control-like profile show greater improvement than patients with abnormal topology. In the rest of this review article, we provide evidence for these emerging trends, from both the extant literature and our own ongoing work in this area.
Several Bilateral Regions Constitute a Network Engaged in Language Recovery
One well-substantiated finding in the literature examining language recovery in aphasia is that several bilateral regions are activated to subserve language recovery. As shown in Figure 2, activation during naming and semantic processing in PWA has been observed in the left supplementary motor area (LSMA; Crosson et al., 2005; Nardo, Holland, Leff, Price, & Crinion, 2017; Saur et al., 2006), left middle frontal gyrus (LMFG; Allendorfer, Kissela, Holland, & Szaflarski, 2012; Cao, Vikingstad, George, Johnson, & Welch, 1999; Fridriksson, 2010; Fridriksson, Bonilha, Baker, Moser, & Rorden, 2010; Sebastian & Kiran, 2011; Szaflarski et al., 2011), left inferior frontal gyrus (LIFG; Cao et al., 1999; Fridriksson, 2010; Léger et al., 2002; Rochon et al., 2010; Saur et al., 2006; Sebastian & Kiran, 2011; Sims et al., 2016; Szaflarski et al., 2011; van Hees, McMahon, Angwin, de Zubicaray, & Copland, 2014; van Oers et al., 2010; Vitali et al., 2007), left precentral gyrus (LPCG; Fridriksson, 2010; Nardo et al., 2017), left superior parietal lobule (LSPL; Fridriksson, 2010; Szaflarski, Allendorfer, Banks, Vannest, & Holland, 2013), left inferior parietal lobule (LIPL; Fridriksson, 2010; Raboyeau et al., 2008; Szaflarski et al., 2011), left angular gyrus (LAG; Crosson et al., 2005; van Hees, McMahon, Angwin, de Zubicaray, & Copland, 2014), left supramarginal gyrus (LSMG; Crosson et al., 2005; Léger et al., 2002; Rochon et al., 2010; van Hees, McMahon, Angwin, de Zubicaray, & Copland, 2014; Vitali et al., 2007), left superior temporal gyrus (LSTG; Crosson et al., 2005; Fernandez et al., 2004), left middle temporal gyrus (LMTG; Fernandez et al., 2004; Rochon et al., 2010; Sebastian & Kiran, 2011; Szaflarski et al., 2011; Vitali et al., 2007), and left inferior temporal gyrus (LITG; Crosson et al., 2005; Saur et al., 2006; van Hees, McMahon, Angwin, de Zubicaray, & Copland, 2014).
It should be noted that several of the same studies have also found activation in RH regions, including the right supplementary motor area (RSMA; Fridriksson, 2010; Fridriksson, Baker, & Moser, 2009), right precentral gyrus (RPCG; Fridriksson et al., 2009), right angular gyrus (RAG; Gold & Kertesz, 2000; Sims et al., 2016), right superior temporal gyrus (RSTG; Fridriksson et al., 2009; Gold & Kertesz, 2000; Meinzer et al., 2006; Skipper-Kallal, Lacey, Xing, & Turkeltaub, 2017; Vitali et al., 2007), right middle temporal gyrus (RMTG; Fridriksson et al., 2009), right temporal pole (RTP; Fridriksson et al., 2009), and right inferior frontal gyrus (RIFG; Fridriksson et al., 2009; Meinzer et al., 2006; Mohr, Difrancesco, Harrington, Evans, & Pulvermüller, 2014; Nardo et al., 2017; Raboyeau et al., 2008; Skipper-Kallal et al., 2017; Vitali et al., 2007). In addition to these “traditional” language regions in the LH and homologous counterparts, studies have also observed activation in domain-general regions, including the left medial frontal gyrus (LmedFG; Allendorfer et al., 2012; Fridriksson, 2010), right middle frontal gyrus (RMFG; Gold & Kertesz, 2000; Raboyeau et al., 2008), left anterior cingulate (LACC; Allendorfer et al., 2012; Fridriksson et al., 2010; Nardo et al., 2017; Sims et al., 2016), right anterior cingulate (RACC; Nardo et al., 2017; Raboyeau et al., 2008; Sims et al., 2016), left precuneus (LPCUN; Fridriksson, 2010; Vitali et al., 2007; Raboyeau et al., 2008), right precuneus (RPCUN; Raboyeau et al., 2008), left posterior cingulate (LPCC; Raboyeau et al., 2008), and right insula (RINS; Nardo et al., 2017; Raboyeau et al., 2008). As noted in the figure, some of these studies have examined language processing at a single time point, whereas others have examined recovery as a function of treatment (denoted by T). One caveat to be noted is that even studies that have examined language processing and not language recovery explicitly have included patients in the chronic stage and, hence, also document recovered language. What these results collectively suggest is that recovery of language is likely bilateral and more importantly, recovery likely involves a network of regions. Thus, it is not surprising that, across studies, specific regions of activation have varied, but Figure 2 demonstrates some degree of consensus among regions that are typically engaged in language recovery. As to which specific regions are engaged in the network is likely determined by the degree and location of damage in the LH across individual patients and the different functional magnetic resonance imaging (fMRI) tasks that have been implemented across the studies reported above.
While these fMRI studies illustrate that specific regions in both hemispheres are concurrently active during language processing after a stroke, to completely understand how these regions interact with each other, one must look at studies that examine connectivity within these regions. Functional or effective connectivity studies typically examine temporal synchronization between different cortical regions and provide a more detailed view of how regions may interact and exchange information with each other. The notion that disrupted connectivity relates to behavioral deficits and recovery after a stroke has been put forth by several studies that examined connectivity globally (i.e., across the whole brain; Siegel et al., 2016, 2018; Y. Zhu et al., 2017) and/or with respect to specific networks, including the default mode network (DMN; Tuladhar et al., 2013), dorsal attention network (Carter et al., 2010), language network (Nair et al., 2015), and frontoparietal and cingulo-opercular cognitive control networks (Nomura et al., 2010). Notably, all of the aforementioned studies have examined connectivity using resting-state fMRI (rsFMRI) data. With regard to language recovery after a stroke, two studies also examining rsFMRI data have reported abnormal left frontoparietal language/cognitive processing networks (D. Zhu et al., 2014) and decreased local homogeneity (i.e., temporal synchronization) in regions including the left calcarine, left lingual, left superior frontal gyrus (SFG), and left medial superior frontal gyrus (SFG; Yang, Li, Yao, & Chen, 2016). Finally, two studies that have reported abnormal measures of functional connectivity and network topology in PWA in the chronic stage include one that identified six resting-state networks and a semantic network that were abnormal compared to controls (Sandberg, 2017) and one that used task-based fMRI to identify abnormal connectivity between left–right anterior lateral and superior temporal cortices (Warren, Crinion, Lambon Ralph, & Wise, 2009). All these studies suggest that stroke inextricably alters the natural but complex dynamic networks that naturally occur in the brain (Thomas Yeo et al., 2011). While these studies report the presence of dysfunction within these networks, our understanding of how these networks recover and reorganize over the course of recovery is only beginning (Siegel et al., 2018).
Studies that have examined treatment-related changes in network connectivity have provided the best evidence yet of how these networks are disrupted relative to controls prior to treatment and, consequently, undergo further change as a function of the treatment. One of the earliest studies was by Abutalebi, Della Rosa, Tettamanti, Green, and Cappa (2009), who found increased activation and connectivity of the LH language network that was associated with improved naming performance in the trained language in one bilingual patient. Similarly, Vitali et al. (2010) studied changes in connectivity in two patients after a phonological cueing treatment. One patient, who suffered a traumatic brain injury, showed increased connectivity in bilateral inferior frontal gyrus (IFG) triangularis and LSMG after treatment. The other patient, who had a stroke, showed increased connectivity in bilateral supramarginal gyrus (SMG), RIFG triangularis, and RMTG. Another study that found increased connectivity as a function of treatment was done by Marcotte, Perlbarg, Marrelec, Benali, and Ansaldo (2013). In this study, the authors examined the DMN in nine patients before and after a semantic-based intervention. Patients showed increased integration of the posterior subnetwork within the DMN as a function of treatment. Van Hees et al. (2014) examined changes in connectivity in eight patients who received phonological and semantic-based treatments. This study employed rsFMRI data and utilized a measure of the power of the data (amplitude of low-frequency fluctuations [ALFF]) to examine fluctuations in the blood oxygen level–dependent (BOLD) signal within regions and synchronization between regions. Pretreatment ALFF in RMTG was associated with higher outcomes in treatment, whereas posttreatment ALFF in LMTG, SMG, and RIFG correlated with higher treatment outcomes. Furthermore, patients had weaker connectivity between LMTG and superior temporal gyrus (STG) as well as between RIFG and LIFG prior to treatment compared to healthy controls, although these differences were no longer significant after treatment. Additionally, increased connectivity between left superior marginal gyrus and SFG, middle frontal gyrus (MFG), and superior parietal lobule (SPL) was observed in patients after treatment. In a similar study, Kiran, Meier, Kapse, and Glynn (2015) examined task-based fMRI data of picture naming and semantic feature processing in eight patients who received semantic-based naming treatment. While there were differences across patients in terms of how connectivity between regions changed within the network as a function of treatment, LIFG and its connectivity significantly increased in more patients than connections from any other region. Sandberg, Bohland, and Kiran (2015) examined 10 patients who received a semantic-based treatment for word retrieval of abstract and concrete words. Sandberg et al. examined brain networks as they were engaged during abstract word or concrete word processing and found an increase in node degree (i.e., hub-like nature) of the LIFG after treatment in the abstract word network (abstract words were trained for all patients) and an increase in node degree of the left medial SFG, and RIFG triangularis was observed for the untrained concrete word network. Another study using graph theoretical approaches to examine network connectivity examined 12 patients who received an imitation-based treatment. As in the van Hees et al. study, Duncan and Small (2016) examined rsFMRI data before and after treatment and found that behavioral improvements in narrative telling were associated with an increase in modularity within the resting-state network. Finally, a recent treatment study by Tao and Rapp (2019) conducted on 15 patients with spelling and writing deficits found that with improved spelling skills, increased network modularity that was, in turn, driven by increased integration of local networks involved in spelling abilities was observed.
To summarize, these aforementioned studies have all focused on treatment-induced changes in network connectivity, and while several have observed increases in connectivity, some, such as van Hees et al. (2014), have found that treatment induces a trend toward more typical/healthy functional connectivity, and others, such as Sandberg et al. (2015), have found more subtle shifts within the networks after treatment. Nonetheless, all these studies underscore the points that language therapy may impact brain reorganization and that changes occur within a network of regions, instead of simply causing individual regions to increase or decrease in activation.
Spared LH Regions Are Important Components in the Network Engaged in Language Recovery
In the previous section, we reviewed evidence that indicates language recovery is not subserved by one or two regions in the LH or RH but instead encompasses a network of regions in both hemispheres. In this section, we posit that, even though a network of regions in both hemispheres is engaged in recovery, remaining, spared tissue within the LH in particular is critically engaged in language recovery. This premise builds on Heiss and Thiel's (2006) hierarchy, where they suggest that both complete and incomplete recoveries include the engagement of undamaged ipsilesional LH regions. Support for this notion comes from a meta-analysis by Turkeltaub, Messing, Norise, and Hamilton (2011), which included fMRI studies where patients either had sustained damage to the LIFG regions or had spared LIFG regions. The results showed that, in studies and patients where LIFG was not damaged, activation was observed in LIFG pars triangularis, pars orbitalis, and pars opercularis; RIFG anterior pars triangularis; LMTG; RMTG; and LMFG. When LIFG was damaged, the shift to RH homologous regions was distinct, with activation observed in the RPCG (precentral/postcentral gyrus), right dorsal IFG pars opercularis, RIFG pars triangularis, RIFG pars orbitalis, and left anterior insula. Notably, irrespective of whether or not LIFG was damaged, activation was observed in the LMFG and right ventral IFG pars opercularis, a point we will return to later in this review article. Nonetheless, the main observation of the study was that a network of residual spared regions in the LH language network was engaged for language recovery and that homologous regions in the RH were active when the LH, specifically the left inferior frontal cortex, was lesioned.
A similar interpretation was drawn by another study that examined the relationship between residual tissue in the LH, BOLD signal activation during semantic processing tasks, and performance on semantic judgment tasks (Sims et al., 2016). Sims et al. (2016) found that, despite extensive damage to LIFG triangularis and opercularis, greater spared tissue in these regions was associated with higher language performance during the fMRI task. Also, the more the damage to LIFG, LMTG, and LAG/SMG, the higher the BOLD activation in spared bilateral frontal regions, including the SFG, MFG, and anterior cingulate cortex (ACC), even though these regions are not traditionally involved in language processing. Finally, when PWA had extensive damage to the entire LH, posterior RH regions (i.e., RMTG, right supramarginal gyrus [RSMG], RAG) became most engaged.
Slightly unrelated but consonant evidence that underscores the importance of spared LH regions in subserving language recovery comes from recent work in our laboratory that examined the structural integrity of gray and white matter regions in their ability to predict language severity and language treatment outcomes. We should note that, unlike the aforementioned studies, this work does not include fMRI data but examines the relationship between structural integrity and language performance. In this study (Meier, Johnson, Pan, & Kiran, 2019b), 34 patients with LH strokes and chronic aphasia underwent a semantic-based naming intervention. Spared voxels in gray matter regions of interest (ROIs) and fractional anisotropy (FA) from white matter tracts in the LH were calculated. Results showed that a combined gray and white matter estimate of structural integrity predicted both naming severity and language treatment outcomes. Additionally, lesion volume accounted for aphasia severity over the structural integrity of any specific LH tract (i.e., arcuate fasciculus). Notably, RH tract integrity did not account for language severity in aphasia, consistent with other studies (Forkel et al., 2014). By contrast, even after controlling for overall LH damage, significant relationships were found between pretreatment naming abilities and naming treatment outcomes and FA in the left inferior longitudinal and left inferior fronto-occipital fasciculi. Several studies have already documented the importance of the structural integrity of white matter tracts in the preservation of lexical–semantic abilities in patients (Han et al., 2013; Harvey & Schnur, 2015; Ivanova et al., 2016) and their connectivity to other parts of the brain (Gleichgerrcht et al., 2015; Yourganov, Fridriksson, Rorden, Gleichgerrcht, & Bonilha, 2016). As in the Meier et al. (2019b) study, Bonilha, Gleichgerrcht, Nesland, Rorden, and Fridriksson (2016) have found left temporal lobe connectivity to be critical for success in naming treatment and for fibers in the left inferior longitudinal fasciculus to change as a function of naming treatment in patients with chronic aphasia (McKinnon et al., 2017). Along with the other studies mentioned above, these results suggest that white matter integrity, and not just integrity of specific regions in the LH, may better predict language recovery outcomes, including outcomes after treatment. As noted in the beginning of this section, it seems that the undamaged gray matter regions and white matter tracts in the LH are important drivers in the process of language recovery.
As Damage Increases in the LH, Activation Expands to the RH and Domain-General Regions
As noted above, based on Sims et al. (2016), the greater the lesion volume in the LH, the greater the activation in RH regions. Another intriguing finding from this study was that bilateral frontal regions, including the SFG, MFG, and ACC, appear to serve as an assistive network in the case of damage to traditional language regions (i.e., IFG, middle temporal gyrus [MTG], angular gyrus [AG], SMG). Notably, in their meta-analytical review of fMRI studies in aphasia, Turkeltaub et al. (2011) also observed that, irrespective of whether the lesion included LIFG or not, activation was observed in the LMFG for patients and not for controls. They interpreted these findings as increased reliance on domain-general regions that may be engaged in cognitive control, regions that are active for patients but are not required for healthy adults during language processing. The particular role of LMFG and its relation to LIFG and LMTG were examined in two studies by our group using fMRI and effective connectivity in groups of PWA and healthy controls. In the first study (Meier, Kapse, & Kiran, 2016), 13 patients and 10 controls performed a picture naming task within the scanner, and connectivity between LMFG, LIFG, and LMTG was examined. This work utilized dynamic causal modeling (DCM) to examine the effective connectivity between these regions and relate it to lexical–semantic abilities in the scanner. Briefly, DCM (Friston, Harrison, & Penny, 2003) is a hypothesis-driven approach that comprises the following steps: (a) definition of specific hypotheses; (b) identification of effects of interest in the fMRI task; (c) building of a DCM model that includes ROIs and connections between these regions; (d) definition of all plausible models that may explain the effects of interest with differing inputs to the model; (e) using the fMRI data, extraction of volumes of interest that correspond to the ROIs; (f) estimation of Bayesian models for all subjects; (g) determination of model fit using Bayesian model selection; and (h) extraction of inference on parameters and connections (Seghier, Zeidman, Neufeld, Leff, & Price, 2010). The last few steps allow for the selection of the optimal model comprising the ROIs and connections that best fit the data as well as the interpretation of the strength of connectivity between regions. In the Meier et al. (2016) study, models that had input to LIFG best explained the control data, but models with input to LMFG best fit the patient data. Two pertinent observations emerged from this study: First, accuracy on the fMRI picture naming task was significantly related to spared tissue in LIFG and LMFG (not LMTG), and second, the higher the spared tissue in these two regions, the higher the strength of connection between them. These observations highlight the point that LMFG is part of a network with LIFG and influences accurate language processing, particularly in patients.
In a follow-up study (Meier, Johnson, & Kiran, 2018), a larger group of 25 patients and 18 healthy controls performed a semantic feature verification task in the scanner. As in the previous study, the ROIs included the LMFG, LIFG, and LMTG, but in this case, the DCM results revealed that models with input to LMFG best explained both patient and control data. For both groups, the modulation of LMFG to LIFG and LMTG was strong even though the groups differed in the connectivity modulation between LIFG and LMTG. Importantly, the strength of activity in LIFG and its connectivity with LMTG was positively associated with lexical–semantic abilities, underscoring the importance of these regions in successful language processing. Furthermore, these two studies highlight the importance of LMFG in modulating LIFG and LMTG during lexical–semantic processing tasks, particularly in patients. These results are consistent with recent studies that report reliance on domain-general networks during language tasks for patients relative to healthy controls (Geranmayeh, Chau, Wise, Leech, & Hampshire, 2017; Geranmayeh, Leech, & Wise, 2016; Sharp, Turkheimer, Bose, Scott, & Wise, 2010). As previously noted in an earlier section (Several Bilateral Regions Constitute a Network Engaged in Language Recovery), several treatment studies have also reported activation in regions, such as the LMFG, that change after treatment (Fridriksson, 2010; Fridriksson et al., 2010; Kiran et al., 2015; Marcotte et al., 2012; Menke et al., 2009; Sandberg et al., 2015). Importantly, domain-general regions such as the LMFG appear to be engaged when damage to the frontotemporal language network is substantial (Turkeltaub et al., 2011); however, they can also be recruited when frontal and/or temporal regions are spared (Kiran et al., 2015; Meier et al., 2016). All these results suggest that, following damage to traditional language regions such as the LIFG and/or LMTG, domain-general regions such as the LMFG and others (e.g., SFG, SPL) may mediate language recovery in patients and become recruited as part of the “altered” network for recovery. We should note that it is unclear to what extent domain-general regions such as the LMFG are engaged in language processing as a compensatory mechanism versus in language processing even in healthy controls. One piece of evidence that likely contributes to this conundrum is the observation that bilateral MD networks in the frontal and parietal cortices are engaged in cognitively demanding or effortful tasks across diverse domains (Duncan & Owen, 2000; Fedorenko, 2014; Fedorenko, Duncan, & Kanwisher, 2013). Thus, future research will illuminate the precise role of domain-general regions, such as the LMFG, in aphasia recovery.
Another aspect of the altered network for recovery involves the expansion of this network to include homologous RH regions. As noted by Turkeltaub et al. (2011), Sims et al. (2016), and others, large lesions in the LH result in the activation of the RH frontal and temporal regions. A recent article by our group illustrates the complex relationship between the degree of damage to the LH, connectivity in the spared regions of the LH and undamaged RH homologues, and language performance. In this study (Meier, Johnson, Pan, & Kiran, 2019a), 30 PWA and 18 healthy controls performed a semantic feature verification task. Unlike the preceding DCM studies described, a more extensive model with bilateral connections was examined. Specifically, four sets of plausible models were created, which depicted left-lateralized connectivity with minimal damage, bilateral anterior-weighted connectivity with posterior damage, bilateral posterior-weighted connectivity with anterior damage, and right-lateralized connectivity with extensive damage to the LH. While healthy controls demonstrated a preference for the left-lateralized connected models, patients demonstrated a split between that set of models and another set that reflected bilateral posterior-weighted connectivity subsequent to anterior damage. For patients, there were several connections between and within the LH and RH regions that were significantly modulated as compared to primarily within-hemisphere connections in the LH for controls. Specifically, patients—but not controls—significantly recruited additional left, frontal intrahemispheric (LIFG➔LMFG), interhemispheric (LIFGtriangularis➔RIFGtriangularis, right inferior temporal gyrus [RITG]➔LITG), and right intrahemispheric (RIFGtriangularis➔RMTG, RITG➔RIFGtriangularis, RITG➔RMTG) connections. Notably and consistent with other studies reviewed in this section, greater spared tissue in LMFG and LMTG was associated with stronger left intrahemispheric connectivity. Better performance on the fMRI task (according to accuracy and reaction times) was related to smaller overall LH lesion volume and left intrahemispheric connectivity, further confirming the importance of preserved LH regions in recovery.
To summarize, these results further underscore the claim that a network of regions is activated in the service of language recovery, and to the extent that LH regions are spared, they are essential and engaged in language recovery. Furthermore, as damage increases in the LH, activation shifts to the RH and MD or domain-general regions such as the MFG.
Patients With Normal-Like Network Topology Show Greater Improvement Than Patients With Abnormal Topology
A final observation that is emerging from current research is that instantiation of varying degrees of network topology determines the degree of language recovery for individual patients. Briefly, network topology refers to the arrangement of and relationships or connections between regions of the brain. One way to characterize network topology is through the application of graph theoretical methods. In graph theory, a network is represented as a graph composed of nodes (regions) and edges (correlation coefficients representing connections between regions). From these nodes and edges, several properties of the network can be extracted. For example, network integration is the capability to transmit and combine information from regions throughout the network (Rubinov & Sporns, 2010). Integration measures include network strength, which describes the overall network connectivity based on each node's direct connections, and global efficiency, which describes how easily nodes in the network can communicate with each other based on direct and indirect connections. Thus, higher network strength and global efficiency are indicative of a highly connected and integrated network. In contrast, network segregation is the capacity of the network to support local processing within densely interconnected brain regions (Rubinov & Sporns, 2010). The measure of average clustering coefficient describes the extent of clustering within the network, whereas network local efficiency describes the efficiency of communication within local clusters. Thus, higher average clustering coefficients and local efficiency are indicative of a network composed of specialized clusters that efficiently communicate among themselves. A healthy efficient network would presumably be optimally segregated into clusters of highly connected regions (perhaps as indicated by the aforementioned clustering coefficient and local efficiency) and highly integrated (per measures such as global efficiency) to facilitate easy transmission of information between distinct regions and clusters (Rubinov & Sporns, 2010). After damage to one or more nodes or cluster of nodes in the brain, this topographical organization may be disrupted, resulting in reduced activation in surrounding nodes and nodes remote from the site of damage; it may also be altered as nodes distant from the lesion that are not typically engaged in a particular function become abnormally engaged (Fornito, Zalesky, & Breakspear, 2015).
Figure 3 summarizes how fMRI data have been transformed into graph theoretical networks in our own recent work: (a) preprocessing the fMRI time series data to remove any outliers due to movement, (b) denoising the data to remove potential confounds in the BOLD signal, (c) selection of a set of ROIs based on the task of interest and modified to account for patients' lesions, (d) extraction of time course of signal in each of these ROIs, (e) generation of a matrix of pairwise correlations of the ROI time series data, and (f) calculation of a weighted undirected graph from the positive correlations.
Our own work in this topic has centered on network factors that explain language recovery after rehabilitation. Preliminary analyses of data collected in our lab from 2013–2018 show that functional connectivity in 26 patients was reduced compared to healthy controls prior to treatment. However, connectivity in patients increased after treatment, particularly among those who had the best treatment outcomes (see Gilmore, Meier, Johnson, & Kiran, 2018, for a description of the treatment study and patients' differential responsiveness to treatment). Conversely, no significant change in connectivity was found in patients who were less successful in treatment or did not undergo treatment. Analyses of several graph measures revealed no significant differences between healthy controls and responders, indicating that they had relatively normal (i.e., control-like) network integration and segregation. However, nonresponders had significantly lower average strength and global efficiency than healthy controls and responders as well as near–significantly lower clustering coefficient and local efficiency than healthy controls. Not surprisingly, the nonresponders had larger lesions and more severe language deficits than responders, even before treatment. These results suggest that individual differences in treatment outcomes may have been related to differences in functional connectivity patterns and network characteristics, such that patients who improved in treatment had a more normal-like network to begin with and patients who did not improve in treatment had more atypical/dysfunctional networks. Similar interpretations were drawn in the previously mentioned article by Tao and Rapp (2019), which examined spelling deficits and response to a subsequent spelling treatment and found that higher baseline modularity and local within-module integration predicted better responsiveness to treatment.
These observations resonate with our related work on structural predictors (Meier et al., 2019b), where we observed that the integrity of many LH structures (white matter tracts and cortical regions) was related to treatment gains. Combined with the findings that FA in the left inferior longitudinal and left inferior fronto-occipital fasciculi was related to higher treatment gains, these results indicate that treatment responders had smaller lesions and higher pretreatment integrity of white matter tracts.
While the preliminary results outlined above suggest that variability in treatment outcomes among PWA may be related to functional connectivity and network characteristics, a limitation of these analyses was the assumption that the treatment (semantic feature analysis) should be an effective treatment for all participants and that everyone who received the treatment would be likely to improve. However, there is evidence indicating variable outcomes with this treatment and that it may not be effective for all patients (van Hees, Angwin, McMahon, & Copland, 2013; Wambaugh, Mauszycki, Cameron, Wright, & Nessler, 2013). In a study under review, we (Johnson, Meier, Pan, & Kiran, 2019) examined if differences in network topology existed in patients even before they commenced the semantic-based treatment. Thus, this study examined pretreatment global and nodal properties in the 26 patients that predicted patients' treatment outcomes. Results showed that patients with higher pretreatment network strength and, to a lesser extent, global and local efficiency had a better response to treatment than patients with lower network measures. Interestingly, when responders and nonresponders were directly compared to each other, several differences emerged in the network properties within bilateral frontal and parietal regions. Results based on preliminary analyses show that differences between the responders and nonresponders emerged on node strength and global efficiency, such that responders had higher node strength in LACC, RIFGtri, and RIFG opercularis as well as higher global efficiency in LPCUN, right superior frontal gyrus (RSFG), RAG, and bilateral IFG opercularis and MFG. Notably, aside from LIFG opercularis, no differences between the groups were observed in classical language regions, such as the LMTG, or other temporal regions. These results suggest that responders have greater connectivity in domain-general regions that support cognitive functions and may also subserve language recovery, as these patients went on to improve after the language treatment. Importantly, these results also suggest that higher levels of network integration, in particular, may be a biomarker for treatment-related recovery in those with chronic stroke aphasia. While these results are preliminary and only include 26 patients, they set the stage for future work in this area, which can substantiate these findings and identify other biomarkers for language recovery.
We close this review article with another schematic of a proposed mechanism of language recovery that considers differences in network topology. In Figure 4, we suggest that several factors influence the extent of language recovery. Immediately after a stroke, there is extensive dysfunction in the brain impacting not only the language network but also other networks (Golestani, Tymchuk, Demchuk, Goodyear, & VISION-2 Study Group, 2013; Nair et al., 2015; Park et al., 2011; Siegel et al., 2016, 2018; Tuladhar et al., 2013; D. Zhu et al., 2014). These early changes in neural function are associated with behavioral deficits, including cognitive–linguistic impairments (Siegel et al., 2016, 2018; D. Zhu et al., 2014; Y. Zhu et al., 2017). In the early weeks and months postonset, improvements in cognitive–linguistic functions coincide with the reemergence of more normal-like connectivity and network topology (Nair et al., 2015; Siegel et al., 2018; D. Zhu et al., 2014). However, when patients reach the chronic stage of recovery, functional connectivity and network topology remain abnormal relative to healthy controls, and more severe language deficits are associated with more abnormal network measures (Meier et al., 2019a; Sandberg, 2017). When treatment is introduced in the chronic stage, some patients improve considerably more than others in terms of language outcomes. Furthermore, treatment-related behavioral improvement is associated with increased/normalized connectivity and network measures (Duncan & Small, 2016; Marcotte et al., 2013; Sandberg et al., 2015; van Hees, McMahon, Angwin, de Zubicaray, Read, & Copland, 2014). Critically, there appear to be measurable differences in functional connectivity and network topology between patients who improve and patients who do not improve in treatment that can be identified before treatment is initiated (Johnson et al., 2019; Marcotte et al., 2013). Specifically, patients with more control-like connectivity and higher indices of network integration before treatment may have better behavioral outcomes. Moreover, the integration of bilateral regions outside of the canonical left perisylvian language network may be a key determinant of treatment response, as demonstrated in our own recent work. In summary, in the time between stroke onset and the introduction of treatment in the chronic stage, functional connectivity and network topology normalize more in some patients than in others, and the extent of this normalization is an indicator of how they will respond to treatment in the chronic stage. In fact, network measures such as strength and global efficiency may be proxy indices for reorganization that has already occurred by the time a patient reaches the chronic stage of recovery. Thus, patients who have experienced beneficial reorganization of language functions up to the chronic stage (perhaps indicated by more normalization and/or higher strength and global efficiency) will respond well to treatment, whereas those who have experienced maladaptive reorganization (perhaps indicated by less normalization and/or lower strength and global efficiency) will have a less favorable response to treatment.
Conclusion
To conclude, this article expanded on Heiss and Thiel's (2006) hierarchy of language recovery and identified four emerging themes: (a) Several bilateral regions constitute a network engaged in language recovery; (b) spared LH regions are important components in the network engaged in language recovery; (c) as damage increases in the LH, activation expands to the RH and domain-general regions; and (d) patients with efficient, control-like network topology show greater improvement than patients with abnormal topology. We proposed a mechanistic model of language recovery that accounted for individual differences in behavior, network topology, and treatment responsiveness.
While some of these postulations are well accepted and substantiated with evidence, others are preliminary, and more work in this area is needed to elucidate the details. From a clinical standpoint, understanding the individual differences across patients that lead to differential recovery patterns is important for better clinical decisions and delivery services. Ultimately, continued work in this area will lead us to a better understanding of the mechanisms underlying language recovery, biomarkers that influence recovery, and, consequently, more personalized treatment options for individual patients.
Acknowledgments
This article stems from the 2018 Research Symposium at ASHA Convention, which was supported by the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health under Award R13DC003383. Research reported in this publication was supported by the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health under NIDCD/NIH P50DC012283, subproject PI to Swathi Kiran and 1F31DC015940 (PI: Erin Meier). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnote
1
Average clustering coefficient and local efficiency contributed to models that predicted treatment outcomes and were positively related to behavioral improvement, although their independent associations did not reach significance.
References
Abutalebi, J., Della Rosa, P. A., Tettamanti, M., Green, D. W., & Cappa, S. F. (2009). Bilingual aphasia and language control: A follow-up fMRI and intrinsic connectivity study. Brain and Language, 109(2–3), 141–156.
Allendorfer, J. B., Kissela, B. M., Holland, S. K., & Szaflarski, J. P. (2012). Different patterns of language activation in post-stroke aphasia are detected by overt and covert versions of the verb generation fMRI task. Medical Science Monitor, 18(3), CR135–CR147.
Bonilha, L., Gleichgerrcht, E., Nesland, T., Rorden, C., & Fridriksson, J. (2016). Success of anomia treatment in aphasia is associated with preserved architecture of global and left temporal lobe structural networks. Neurorehabilitation and Neural Repair, 30(3), 266–279.
Cao, Y., Vikingstad, E. M., George, K. P., Johnson, A. F., & Welch, K. M. (1999). Cortical language activation in stroke patients recovering from aphasia with functional MRI. Stroke, 30(11), 2331–2340.
Carter, A. R., Astafiev, S. V., Lang, C. E., Connor, L. T., Rengachary, J., Strube, M. J., … Corbetta, M. (2010). Resting interhemispheric functional magnetic resonance imaging connectivity predicts performance after stroke. Annals of Neurology, 67(3), 365–375.
Crosson, B., Moore, A. B., Gopinath, K., White, K. D., Wierenga, C. E., Gaiefsky, M. E., … Gonzalez Rothi, L. J. (2005). Role of the right and left hemispheres in recovery of function during treatment of intention in aphasia. Journal of Cognitive Neuroscience, 17(3), 392–406.
Duncan, E. S., & Small, S. L. (2016). Increased modularity of resting state networks supports improved narrative production in aphasia recovery. Brain Connectivity, 6(7).
Duncan, J., & Owen, A. M. (2000). Common regions of the human frontal lobe recruited by diverse cognitive demands. Trends in Neurosciences, 23(10), 475–483.
Fedorenko, E. (2014). The role of domain-general cognitive control in language comprehension. Frontiers in Psychology, 5, 335.
Fedorenko, E., Duncan, J., & Kanwisher, N. (2013). Broad domain generality in focal regions of frontal and parietal cortex. Proceedings of the National Academy of Sciences of the United States of America, 110(41), 16616–16621.
Fernandez, B., Cardebat, D., Demonet, J.-F., Joseph, P. A., Mazaux, J.-M., Barat, M., & Allard, M. (2004). Functional MRI follow-up study of language processes in healthy subjects and during recovery in a case of aphasia. Stroke, 35(9), 2171–2176.
Forkel, S. J., Thiebaut de Schotten, M., Dell'Acqua, F., Kalra, L., Murphy, D. G., Williams, S. C., & Catani, M. (2014). Anatomical predictors of aphasia recovery: A tractography study of bilateral perisylvian language networks. Brain, 137(7), 2027–2039.
Fornito, A., Zalesky, A., & Breakspear, M. (2015). The connectomics of brain disorders. Nature Reviews Neuroscience, 16(3), 159–172.
Fridriksson, J. (2010). Preservation and modulation of specific left hemisphere regions is vital for treated recovery from anomia in stroke. The Journal of Neuroscience, 30(35), 11558–11564.
Fridriksson, J., Baker, J. M., & Moser, D. (2009). Cortical mapping of naming errors in aphasia. Human Brain Mapping, 30(8), 2487–2498.
Fridriksson, J., Bonilha, L., Baker, J. M., Moser, D., & Rorden, C. (2010). Activity in preserved left hemisphere regions predicts anomia severity in aphasia. Cerebral Cortex, 20(5), 1013–1019.
Friston, K. J., Harrison, L., & Penny, W. (2003). Dynamic causal modelling. NeuroImage, 19(4), 1273–1302.
Geranmayeh, F., Chau, T. W., Wise, R. J. S., Leech, R., & Hampshire, A. (2017). Domain-general subregions of the medial prefrontal cortex contribute to recovery of language after stroke. Brain, 140(7), 1947–1958.
Geranmayeh, F., Leech, R., & Wise, R. J. S. (2016). Network dysfunction predicts speech production after left hemisphere stroke. Neurology, 86(14), 1296–1305.
Gleichgerrcht, E., Kocher, M., Nesland, T., Rorden, C., Fridriksson, J., & Bonilha, L. (2015). Preservation of structural brain network hubs is associated with less severe post-stroke aphasia. Restorative Neurology and Neuroscience, 34(1), 19–28.
Gilmore, N., Meier, E. L., Johnson, J. P., & Kiran, S. (2018). Typicality-based semantic treatment for anomia results in multiple levels of generalisation. Neuropsychological Rehabilitation, 1–27.
Gold, B. T., & Kertesz, A. (2000). Right hemisphere semantic processing of visual words in an aphasic patient: An fMRI study. Brain and Language, 73(3), 456–465.
Golestani, A.-M., Tymchuk, S., Demchuk, A., Goodyear, B. G., & VISION-2 Study Group. (2013). Longitudinal evaluation of resting-state fMRI after acute stroke with hemiparesis. Neurorehabilitation and Neural Repair, 27(2), 153–163.
Han, Z., Ma, Y., Gong, G., He, Y., Caramazza, A., & Bi, Y. (2013). White matter structural connectivity underlying semantic processing: Evidence from brain damaged patients. Brain, 136(10), 2952–2965.
Harvey, D. Y., & Schnur, T. T. (2015). Distinct loci of lexical and semantic access deficits in aphasia: Evidence from voxel-based lesion-symptom mapping and diffusion tensor imaging. Cortex, 67, 37–58.
Heiss, W. D., & Thiel, A. (2006). A proposed regional hierarchy in recovery of post-stroke aphasia. Brain and Language, 98(1), 118–123.
Hillis, A. E., & Heidler, J. (2002). Mechanisms of early aphasia recovery. Aphasiology, 16(9), 885–895.
Ivanova, M. V., Isaev, D. Y., Dragoy, O. V., Akinina, Y. S., Petrushevskiy, A. G., Fedina, O. N., … Dronkers, N. F. (2016). Diffusion-tensor imaging of major white matter tracts and their role in language processing in aphasia. Cortex, 85, 165–181.
Johnson, J. P., Meier, E. L., Pan, Y., & Kiran, S. (2019). Pre-treatment graph measures of a functional semantic network are associated with naming therapy outcomes in chronic aphasia. Manuscript submitted for publication .
Kiran, S., Meier, E. L., Kapse, K. J., & Glynn, P. A. (2015). Changes in task-based effective connectivity in language networks following rehabilitation in post-stroke patients with aphasia. Frontiers in Human Neuroscience, 9, 316.
Kiran, S., & Thompson, C. K. (2019). Neuroplasticity of language networks in aphasia: Advances, updates and future challenges. Frontiers in Neurology: Stroke, 10, 295.
Léger, A., Démonet, J. F., Ruff, S., Aithamon, B., Touyeras, B., Puel, M., … Cardebat, D. (2002). Neural substrates of spoken language rehabilitation in an aphasic patient: An fMRI study. NeuroImage, 17(1), 174–183.
Marcotte, K., Adrover-Roig, D., Damien, B., de Préaumont, M., Généreux, S., Hubert, M., & Ansaldo, A. I. (2012). Therapy-induced neuroplasticity in chronic aphasia. Neuropsychologia, 50(8), 1776–1786.
Marcotte, K., Perlbarg, V., Marrelec, G., Benali, H., & Ansaldo, A. I. (2013). Default-mode network functional connectivity in aphasia: Therapy-induced neuroplasticity. Brain and Language, 124(1), 45–55.
McKinnon, E. T., Fridriksson, J., Glenn, G. R., Jensen, J. H., Helpern, J. A., Basilakos, A., … Bonilha, L. (2017). Structural plasticity of the ventral stream and aphasia recovery. Annals of Neurology, 82(1), 147–151.
Meier, E. L., Johnson, J. P., & Kiran, S. (2018). Left frontotemporal effective connectivity during semantic feature judgments in patients with chronic aphasia and age-matched healthy controls. Cortex, 108, 173–192.
Meier, E. L., Johnson, J. P., Pan, Y., & Kiran, S. (2019a). A lesion and connectivity-based hierarchical model of chronic aphasia recovery dissociates patients and healthy controls. NeuroImage: Clinical, 23, 101919.
Meier, E. L., Johnson, J. P., Pan, Y., & Kiran, S. (2019b). The utility of lesion classification in predicting language and treatment outcomes in chronic stroke-induced aphasia. Brain Imaging and Behavior, 1–16.
Meier, E. L., Kapse, K. J., & Kiran, S. (2016). The relationship between frontotemporal effective connectivity during picture naming, behavior, and preserved cortical tissue in chronic aphasia. Frontiers in Human Neuroscience, 10, 109.
Meinzer, M., Flaisch, T., Obleser, J., Assadollahi, R., Djundja, D., Barthel, G., & Rockstroh, B. (2006). Brain regions essential for improved lexical access in an aged aphasic patient: A case report. BMC Neurology, 6, 28.
Menke, R., Meinzer, M., Kugel, H., Deppe, M., Baumgärtner, A., Schiffbauer, H., … Breitenstein, C. (2009). Imaging short- and long-term training success in chronic aphasia. BMC Neuroscience, 10, 118.
Mijalkov, M., Kakaei, E., Pereira, J. B., Westman, E., & Volpe, G. (2017). BRAPH: A graph theory software for the analysis of brain connectivity. PLOS ONE, 12(8), e0178798.
Mohr, B., Difrancesco, S., Harrington, K., Evans, S., & Pulvermüller, F. (2014). Changes of right-hemispheric activation after constraint-induced, intensive language action therapy in chronic aphasia: fMRI evidence from auditory semantic processing. Frontiers in Human Neuroscience, 8, 919.
Nair, V. A., Young, B. M., La, C., Reiter, P., Nadkarni, T. N., Song, J., … Prabhakaran, V. (2015). Functional connectivity changes in the language network during stroke recovery. Annals of Clinical and Translational Neurology, 2(2), 185–195.
Nardo, D., Holland, R., Leff, A. P., Price, C. J., & Crinion, J. T. (2017). Less is more: Neural mechanisms underlying anomia treatment in chronic aphasic patients. Brain, 140(11), 3039–3054.
Nomura, E. M., Gratton, C., Visser, R. M., Kayser, A., Perez, F., & D'Esposito, M. (2010). Double dissociation of two cognitive control networks in patients with focal brain lesions. Proceedings of the National Academy of Sciences of the United States of America, 107(26), 12017–12022.
Park, C.-H., Chang, W. H., Ohn, S. H., Kim, S. T., Bang, O. Y., Pascual-Leone, A., & Kim, Y.-H. (2011). Longitudinal changes of resting-state functional connectivity during motor recovery after stroke. Stroke, 42(5), 1357–1362.
Raboyeau, G., De Boissezon, X., Marie, N., Balduyck, S., Puel, M., Bézy, C., … Cardebat, D. (2008). Right hemisphere activation in recovery from aphasia: Lesion effect or function recruitment? Neurology, 70(4), 290–298.
Rochon, E., Leonard, C., Burianova, H., Laird, L., Soros, P., Graham, S., & Grady, C. (2010). Neural changes after phonological treatment for anomia: An fMRI study. Brain and Language, 114(3), 164–179.
Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural Neurology, 12(4), 191–200.
Rubinov, M., & Sporns, O. (2010). Complex network measures of brain connectivity: Uses and interpretations. NeuroImage, 52(3), 1059–1069.
Sandberg, C. W. (2017). Hypoconnectivity of resting-state networks in persons with aphasia compared with healthy age-matched adults. Frontiers in Human Neuroscience, 11, 91.
Sandberg, C. W., Bohland, J. W., & Kiran, S. (2015). Changes in functional connectivity related to direct training and generalization effects of a word finding treatment in chronic aphasia. Brain and Language, 150, 103–116.
Saur, D., Lange, R., Baumgaertner, A., Schraknepper, V., Willmes, K., Rijntjes, M., & Weiller, C. (2006). Dynamics of language reorganization after stroke. Brain, 129(6), 1371–1384.
Sebastian, R., & Kiran, S. (2011). Task-modulated neural activation patterns in chronic stroke patients with aphasia. Aphasiology, 25(8), 927–951.
Seghier, M. L., Zeidman, P., Neufeld, N. H., Leff, A. P., & Price, C. J. (2010). Identifying abnormal connectivity in patients using dynamic causal modeling of fMRI responses. Frontiers in Systems Neuroscience, 4, 142.
Sharp, D. J., Turkheimer, F. E., Bose, S. K., Scott, S. K., & Wise, R. J. (2010). Increased frontoparietal integration after stroke and cognitive recovery. Annals of Neurology, 68(5), 753–756.
Siegel, J. S., Ramsey, L. E., Snyder, A. Z., Metcalf, N. V., Chacko, R. V., Weinberger, K., … Corbetta, M. (2016). Disruptions of network connectivity predict impairment in multiple behavioral domains after stroke. Proceedings of the National Academy of Sciences of the United States of America, 113(30), E4367–E4376.
Siegel, J. S., Seitzman, B. A., Ramsey, L. E., Ortega, M., Gordon, E. M., Dosenbach, N. U. F., … Corbetta, M. (2018). Re-emergence of modular brain networks in stroke recovery. Cortex, 101, 44–59.
Sims, J., Kapse, K., Glynn, P., Sandberg, C., Tripodis, Y., & Kiran, S. (2016). The relationships between the amount of spared tissue, percent signal change, and accuracy in semantic processing in aphasia. Neuropsychologia, 84, 113–126.
Skipper-Kallal, L. M., Lacey, E. H., Xing, S., & Turkeltaub, P. E. (2017). Right hemisphere remapping of naming functions depends on lesion size and location in poststroke aphasia. Neural Plasticity, 2017, 8740353.
Szaflarski, J. P., Allendorfer, J. B., Banks, C., Vannest, J., & Holland, S. K. (2013). Recovered vs. not-recovered from post-stroke aphasia: The contributions from the dominant and non-dominant hemispheres. Restorative Neurology and Neuroscience, 31(4), 347–360.
Szaflarski, J. P., Eaton, K., Ball, A. L., Banks, C., Vannest, J., Allendorfer, J. B., … Holland, S. K. (2011). Poststroke aphasia recovery assessed with functional magnetic resonance imaging and a picture identification task. Journal of Stroke & Cerebrovascular Diseases, 20(4), 336–345.
Tao, Y., & Rapp, B. (2019). The effects of lesion and treatment-related recovery on functional network modularity in post-stroke dysgraphia. NeuroImage: Clinical, 23, 101865.
Thomas Yeo, B. T., Krienen, F. M., Sepulcre, J., Sabuncu, M. R., Lashkari, D., Hollinshead, M., … Buckner, R. L. (2011). The organization of the human cerebral cortex estimated by intrinsic functional connectivity. Journal of Neurophysiology, 106(3), 1125–1165.
Tuladhar, A. M., Snaphaan, L., Shumskaya, E., Rijpkema, M., Fernandez, G., Norris, D. G., & de Leeuw, F.-E. (2013). Default mode network connectivity in stroke patients. PLOS ONE, 8(6), e66556.
Turkeltaub, P. E., Messing, S., Norise, C., & Hamilton, R. H. (2011). Are networks for residual language function and recovery consistent across aphasic patients? Neurology, 76(20), 1726–1734.
van Hees, S., Angwin, A., McMahon, K., & Copland, D. (2013). A comparison of semantic feature analysis and phonological components analysis for the treatment of naming impairments in aphasia. Neuropsychological Rehabilitation, 23(1), 102–132.
van Hees, S., McMahon, K., Angwin, A., de Zubicaray, G., & Copland, D. A. (2014). Neural activity associated with semantic versus phonological anomia treatments in aphasia. Brain and Language, 129, 47–57.
van Hees, S., McMahon, K., Angwin, A., de Zubicaray, G., Read, S., & Copland, D. A. (2014). A functional MRI study of the relationship between naming treatment outcomes and resting state functional connectivity in post-stroke aphasia. Human Brain Mapping, 35(8), 3919–3931.
van Oers, C. A., Vink, M., van Zandvoort, M. J., van der Worp, H. B., de Haan, E. H., Kappelle, L. J., … Dijkhuizen, R. M. (2010). Contribution of the left and right inferior frontal gyrus in recovery from aphasia. A functional MRI study in stroke patients with preserved hemodynamic responsiveness. NeuroImage, 49(1), 885–893.
Vitali, P., Abutalebi, J., Tettamanti, M., Danna, M., Ansaldo, A.-I., Perani, D., … Cappa, S. F. (2007). Training-induced brain remapping in chronic aphasia: A pilot study. Neurorehabiltation and Neural Repair, 21(2), 152–160.
Vitali, P., Tettamanti, M., Abutalebi, J., Ansaldo, A.-I., Perani, D., Cappa, S. F., & Joanette, Y. (2010). Generalization of the effects of phonological training for anomia using structural equation modelling: A multiple single-case study. Neurocase, 16(2), 93–105.
Wambaugh, J. L., Mauszycki, S., Cameron, R., Wright, S., & Nessler, C. (2013). Semantic feature analysis: Incorporating typicality treatment and mediating strategy training to promote generalization. American Journal of Speech-Language Pathology, 22(2), S334–S369.
Warren, J. E., Crinion, J. T., Lambon Ralph, M. A., & Wise, R. J. (2009). Anterior temporal lobe connectivity correlates with functional outcome after aphasic stroke. Brain, 132(12), 3428–3442.
Whitfield-Gabrieli, S., & Nieto-Castanon, A. (2012). CONN: A functional connectivity toolbox for correlated and anticorrelated brain networks. Brain Connectivity, 2(3), 125–141.
Yang, M., Li, J., Yao, D., & Chen, H. (2016). Disrupted intrinsic local synchronization in poststroke aphasia. Medicine, 95(11), e3101.
Yourganov, G., Fridriksson, J., Rorden, C., Gleichgerrcht, E., & Bonilha, L. (2016). Multivariate connectome-based symptom mapping in post-stroke patients: Networks supporting language and speech. The Journal of Neuroscience, 36(25), 6668–6679.
Zhu, D., Chang, J., Freeman, S., Tan, Z., Xiao, J., Gao, Y., & Kong, J. (2014). Changes of functional connectivity in the left frontoparietal network following aphasic stroke. Frontiers in Behavioral Neuroscience, 8, 167.
Zhu, Y., Bai, L., Liang, P., Kang, S., Gao, H., & Yang, H. (2017). Disrupted brain connectivity networks in acute ischemic stroke patients. Brain Imaging and Behavior, 11(2), 444–453.
Information & Authors
Information
Published In
Journal of Speech, Language, and Hearing Research
Volume 62 • Number 11 • November 2019
Pages: 3973-3985
PubMed: 31756154
Copyright
Copyright © 2019 The Authors. Image (CCBY-NC.png) is missing or otherwise invalid.This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
History
- Received: Feb 4, 2019
- Revised: Jun 21, 2019
- Accepted: Aug 27, 2019
- Published online: Nov 22, 2019
- Published in issue: Nov 22, 2019
Authors
Metrics & Citations
Metrics
Article Metrics
View all metricsCitations
Download article citation data for:
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
Citing Literature
- Trevor A. Low, Kevin Lindland, Adam Kirton, Helen L. Carlson, Ashley D. Harris, Bradley G. Goodyear, Oury Monchi, Michael D. Hill, Miranda L. Rose, Sean P. Dukelow, Transcranial Magnetic Stimulation Combined With Multimodality Aphasia Therapy for Chronic Poststroke Aphasia, Neurology, 10.1212/WNL.0000000000213424, 104, 6, (2025).
- Manuel Jose Marte, Marissa Russell-Meill, Nicole Carvalho, Swathi Kiran, Charting the Course of Aphasia Recovery: Factors, Trajectories, and Outcomes, Annual Review of Linguistics, 10.1146/annurev-linguistics-011724-121245, 11, 1, (111-136), (2025).
- Veronika Vadinova, Sonia L E Brownsett, Kimberley L Garden, Tracy Roxbury, Katherine O’Brien, David A Copland, Katie L McMahon, Aleksi J Sihvonen, Early subacute frontal callosal microstructure and language outcomes after stroke, Brain Communications, 10.1093/braincomms/fcae370, 7, 1, (2025).
- Abraham Olufemi Asuku, Emmanuel Sunday Arinde, Salam Babatunde Saliu, Toheeb O. Oyerinde, Tobiloba Samuel Olajide, Hezekiah Omotayo Afolabi, Afeez Kehinde, Neuroplasticity, Advancing Medical Research Through Neuroscience, 10.4018/979-8-3693-5464-3.ch005, (95-118), (2025).
- Swathi Kiran, Maria Varkanitsa, Anne Billot, Erin Carpenter, Isaac Falconer, Individual differences in neuroplastic recovery, Encyclopedia of the Human Brain, 10.1016/B978-0-12-820480-1.00206-0, (758-774), (2025).
- Audra W. Yetter, A Psycholinguistic Approach to the Management of Aphasia in a Case of Left Parietal Stroke, Perspectives of the ASHA Special Interest Groups, 10.1044/2024_PERSP-24-00007, 10, 2, (358-375), (2024).
- Melissa D. Stockbridge, Zafer Keser, Leonardo Bonilha, Argye E. Hillis, Microstructural properties in subacute aphasia: concurrent and prospective relationships underpinning recovery, Brain Structure and Function, 10.1007/s00429-024-02826-z, 229, 9, (2207-2217), (2024).
- Alex Teghipco, Roger Newman-Norlund, Julius Fridriksson, Christopher Rorden, Leonardo Bonilha, Distinct brain morphometry patterns revealed by deep learning improve prediction of post-stroke aphasia severity, Communications Medicine, 10.1038/s43856-024-00541-8, 4, 1, (2024).
- Aleksi J. Sihvonen, Anni Pitkäniemi, Sini-Tuuli Siponkoski, Linda Kuusela, Noelia Martínez-Molina, Sari Laitinen, Essi-Reetta Särkämö, Johanna Pekkola, Susanna Melkas, Gottfried Schlaug, Viljami Sairanen, Teppo Särkämö, Structural Neuroplasticity Effects of Singing in Chronic Aphasia, eneuro, 10.1523/ENEURO.0408-23.2024, 11, 5, (ENEURO.0408-23.2024), (2024).
- Victoria Tilton-Bolowsky, Melissa D. Stockbridge, Argye E. Hillis, Remapping and Reconnecting the Language Network after Stroke, Brain Sciences, 10.3390/brainsci14050419, 14, 5, (419), (2024).
- See more
View Options
View options
Sign In Options
ASHA member? If so, log in with your ASHA website credentials for full access.
Member Login
