The Flexible Brain: (Re-)shaping Adaptation in Semantic Cognition

Project Information

FLEXBRAIN

Grant agreement ID: 101043747

DOI

10.3030/101043747

EC signature date 31 May 2022

Start date 1 January 2023

End date 31 December 2027

Funded under

• European Research Council (ERC)

Total cost

€ 1 999 750,00

EU contribution

€ 1 999 750,00

Coordinated by

UNIVERSITAET LEIPZIG
Germany

Periodic Reporting for period 1 - FLEXBRAIN (The Flexible Brain: (Re-)shaping Adaptation in Semantic Cognition)

Reporting period: 2023-01-01 to 2025-06-30

Summary of the context and overall objectives of the project

The human brain is flexible. Neural networks adapt to cognitive demands by flexibly recruiting different regions and connections. Flexible network adaptation enables cognitive functions such as semantic cognition: the ability to use, manipulate, and generalize knowledge. When key nodes suffer damage, networks can adapt to recover function. Yet, brain lesions often severely impair semantic cognition. How the semantic network adapts to lesions is poorly understood. My hypothesis is that disruption of the semantic network can be compensated for by recruitment of domain-general networks. This notion is based on findings that disruption of semantic nodes inhibits semantic activity but increases activity in domain-general nodes. Yet, the behavioral relevance of domain-general recruitment is unclear. Compensation means that behavior can be preserved as other nodes work harder. Can domain-general networks effectively compensate for disruption of specialized nodes? Is this a common principle of flexible adaptation in the healthy young, aging, and lesioned brain? Unprecedented inhibitory and facilitatory neurostimulation will be used to unbalance and rebalance network adaptation in semantic cognition. Importantly, a novel network stimulation approach will target multiple nodes simultaneously. I ask three questions. (i) Can domain-general networks compensate for semantic network disruption? (ii) Is domain-general recruitment in the aging brain adaptive? (iii) Do domain-general networks drive flexible adaptation to lesions? Perturbing young brains will elucidate the relevance of network adaptation. Perturbing aging brains will probe compensatory reorganization. Facilitating lesioned brains will reshape flexible adaptation. These goals will be addressed in three complementary work packages that combine neurostimulation over single or several key areas with simultaneous functional neuroimaging during semantic and control tasks. Benefitting from my strong neurostimulation experience, we will elucidate the way the brain compensates for disruption. The potential impact of the project on current conceptions of brain plasticity, and for rehabilitative medicine in particular, is immense.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

In the first reporting period, we started all three work packages (WPs), with a focus on WP 1 and establishing the concurrent TMS-fMRI setup. After some technical pilots, we were able to run the first study: a concurrent TMS-fMRI experiment on semantic processing. Given that no study to date has demonstrated the impact of concurrent TMS during fMRI on the semantic system, we first ran a pilot outside the scanner to test the impact of the employed TMS protocol. To account for the frequently reported large individual variability in response to TMS (Turker et al., NeuroImage 2023), we changed our initial paradigm and adopted a semantic matching task with an individual staircase procedure to adjust individual difficulty. We successfully implemented this paradigm and ran a first concurrent TMS-fMRI study with 25 healthy young participants. Results show that TMS significantly increased task-related activity in temporal and frontal semantic areas and in domain-general regions.
Moreover, we employed a behavioral TMS paradigm to probe the relevance of language-related and domain-general areas for semantic and executive control processes. We perturbed either the left IFG (semantic area) or pre-SMA (domain-general area) or both before healthy young volunteers performed semantic fluency and figural fluency tasks (publication in preparation). Perturbation of both areas affected both tasks, arguing for a domain-general contribution of both areas to fluency tasks. We also provided evidence for compensation between both areas. These findings complement and inform our TMS-fMRI studies. We also employed a novel machine learning-based cluster and switching analysis to elucidate the impact of TMS on semantic category switching. These results are relevant for informing WP 2.

For WP 2, we made a step forward in characterizing network interactions. We demonstrated that facilitatory TMS over the pre-SMA induced large-scale network interaction changes across the whole brain (Martin et al., Brain Stimulation 2023). The strength of the modulation between domain-general networks was associated with more efficient semantic performance. Using neuroimaging, we could also show how large-scale network interactions during semantic processing change as we age (Rysop et al. Neurobiology of Aging, in press).

In WP 3, we ran a longitudinal effective connectivity analysis on a data set of stroke patients. We identified three principles of successful language recovery after stroke: First, recruitment of domain-general regions supports language areas by increasing their facilitatory influence on perilesional language areas already during the first days after stroke, which is associated with favourable long-term language recovery. Second, the increase in the interaction between language areas changes over the time course of recovery. Third, early recruitment of the lesion homologue in the right prefrontal cortex is beneficial. These results (Jiang et al., Brain 2025) substantially advance the current knowledge of language recovery.

I consider the following five publications the most significant achievements:

1. Jiang Z, Kuhnke P, Stockert A, Wawrzyniak M, Halai A, Saur D & Hartwigsen G (in press). Dynamic reorganization of task-related network interactions in post-stroke aphasia recovery. Brain. https://academic.oup.com/brain/advance-article/doi/10.1093/brain/awaf036/7990878

2. Turker S, Kuhnke P, Eickhoff SB, Caspers S & Hartwigsen G (2023). Cortical, subcortical, and cerebellar contributions to language processing: A meta-analytic review of 403 neuroimaging experiments. Psychological Bulletin 149, 699-723. https://doi.org/10.1037/bul0000403

3. Numssen O, Kuhnke P, Weise K & Hartwigsen G (2024). Electric field-based dosing for TMS. Imaging Neuroscience 2: 1–12. https://doi.org/10.1162/imag_a_00106

4. Schuler A-L & Hartwigsen G (2025). The potential of interleaved TMS-fMRI for linking stimulation-induced changes in task-related activity with behavioral modulations. Brain Stimulation 18(1): 37–51. https://doi.org/10.1016/j.brs.2024.12.1190

5. Williams K, Numssen O, Guerra, JD, Kopal J, Bzdok D & Hartwigsen G (2024). Inhibition of the inferior parietal lobe triggers state-dependent network adaptations. Heliyon 10 (21): e39735. https://doi.org/10.1016/j.heliyon.2024.e39735

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The most relevant breakthrough for the field is our publication on the network interactions between domain-general areas and language regions during language recovery after stroke (Jiang et al., Brain 2025). These findings support the general idea that areas of the multiple-demand network actively support language recovery after stroke, supporting my framework of domain-general and domain-specific network interactions. Yet, these results significantly extend previous work and our hypotheses on network reorganization, demonstrating that very early on, already in the first days after stroke, domain-general areas are exhibiting a facilitatory influence on language areas, which is associated with less acute impairment and favourable language recovery over the time course of recovery. As a novel and unexpected finding, our results show that such task-related influences precede task-related upregulation of activation in multiple-demand areas.