Russian Researchers Unveil Mechanism Underlying Language Processing Disruptions in Epilepsy Patients

Researchers at HSE University and the Pirogov National Medical and Surgical Centre have examined alterations induced by epilepsy in the language-related neural network within the brain. Using graph-based analysis, the researchers studied fMRI data from 28 patients and found that in epilepsy, both hemispheres of the brain become activated during language processing and short connections form between the hemispheres, while long connections within one hemisphere are disrupted. The study has been published in Epilepsy&Behavior.

Epilepsy affects approximately 1% of the global population. It is a chronic brain disorder that commonly manifests in the temporal lobe. However, epilepsy is not confined to a single brain region; rather, this condition impacts neural networks throughout the brain. Traditionally, the neural network responsible for language processing is believed to be located in the left hemisphere. In individuals with epilepsy, the function of this neural network is disrupted, leading to difficulties with language processing tasks or outright failure to perform them.

Functional magnetic resonance imaging (fMRI) is a standard method employed to investigate changes occurring in the neural networks of patients with epilepsy. Brain scans reveal that in epilepsy, additional regions of the brain are activated in both hemispheres, rather than solely in the left hemisphere where the language function is primarily localised. However, until recently, there was no reliable information available on how these regions interconnect and interact.

Researchers at the HSE Centre for Language and Brain, in collaboration with colleagues at the Pirogov National Medical and Surgical Centre, have developed and applied an unprecedented integrated approach to analyse the language-related neural network in patients with epilepsy. This approach combines mathematical graph theory with fMRI data obtained from subjects performing sentence-completion tasks.

While these two methods have previously been employed independently to investigate brain activity, we are the first to propose their combined use for analysing the language-related network in patients with epilepsy. Only by combining both approaches can we reveal the complete picture. Through fMRI, we measure the activity of brain regions involved in information processing and generation. Observing subjects as they engage in the language task—specifically, a sentence completion task in our case—enables us to understand the interconnectedness of the activated brain regions. Subsequently, graph theory allows us to describe factors such as the strength or weakness of connections between specific regions, their interconnectedness, and division into subnets.

Viktor Karpychev
Co-author of the study, Research Fellow, HSE Centre for Language and Brain

The study involved 28 patients with drug-resistant epilepsy, and 19 healthy volunteers as the control group. During the fMRI session, the subjects engaged in a sentence completion task. They viewed a sentence on the screen with the end omitted, eg 'Our intelligent neighbour has read...' and were instructed to complete the sentence in a manner that was both grammatically and semantically correct.

The study revealed that in healthy individuals, language activity is concentrated in the left hemisphere and is characterised by long intra-hemispheric connections. In patients with epilepsy, however, the language-related network undergoes reorganisation due to their condition, resulting in difficulties with language and speech. Thus, the study subjects with epilepsy performed poorly during the sentence completion test. The application of graph theory helped the researchers uncover that in epilepsy, long connections within the left hemisphere are disrupted, while short modules connecting the left and right hemispheres are formed instead.

Thanks to this new integrated approach, for the first time, we were able to observe changes in the language-related network of patients with epilepsy not only structurally but also functionally. This understanding has shed light on why they encounter more challenges in language processing tasks. We intend to continue investigating this area to explore how the language-related network interacts with other neural networks, both in healthy individuals and in patients with epilepsy. Understanding how the brain reorganises itself under the influence of various diseases can lead to significant improvements in patients' quality of life by allowing us to compensate for the effects of the disease.

Svetlana Malyutina
Co-author of the study, Deputy Director, HSE Centre for Language and Brain

IQ

February 27