Abstracts

Rationale: Several studies based on resting-state functional MRI have reported that epilepsy is associated with brain network reorganization. Impaired memory and executive functions are associated with disrupted regional connectivity. However, cognitive processes are dynamic. Brain connectivity simultaneously and continuously changes throughout the performance of cognitive tasks. This is the first study to use functional near-infrared spectroscopy (fNIRS) to determine the means by which epileptogenic side influences language-associated networks. Methods: We recruited 11 left fronto-temporal epilepsy patients (L-FTE), 11 right fronto-temporal epilepsy patients (R-FTE), and 22 age- and sex-matched controls. Signals from 52-channel fNIRS placed over fronto-temporal scalp were recorded, while the subject was undertaking the categorical fluency test (CFT) and letter fluency test (LFT). We chose the top 10% strongest connections in each three groups, to identify their differences and to find out the relationship between the network topology and the clinical conditions. Results: When performing CFT, the 10% strongest connections showed R-FTE had higher node degree (p = 0.002), network density(p=0.001), and weight strength (p = 0.010) in inter-hemispheric connections compared with controls. L-FTE had decreased right intra-hemispheric path length compared with the controls (p = 0.033). The CFT scores were correlated with the left intra-hemispheric path length (p = 0.030) and inter-hemispheric path length (p = 0.008) only in L-FTE, not in R-FTE. This suggested L-FTE struggled for less efficient inter- and intra- hemispheric connections to compensate the lesion-related functional deficits.When performing LFT, the 10% strongest connections did not have statistical differences between any epilepsy groups and controls. Seizure duration in L-FTE was negatively related to left intra-hemispheric density (p = 0.035), node degree (p = 0.020), and weight strength (p = 0.045). LFT scores in L-FTE were positively correlated with left intra-hemispheric density (p = 0.017), node degree (p = 0.047), weight strength (p = 0.031), while negatively related to right intra-hemispheric weight strength (p = 0.031). LFT scores in R-FTE were positively related to right intra-hemispheric density (p = 0.030), node degree (p = 0.023), weight strength (p = 0.007), while negatively related to left intra-hemispheric density (p = 0.041), node degree (p = 0.026), weight strength (p = 0.034). Seizure duration in L-FTE seems mostly to threaten the left intra-hemispheric connections. LFT performance is related to the preserved network topology ipsilateral to the lesion side. Conclusions: L-FTE and R-FTE differ in terms of task related network reshaping patterns. Bilateral temporal cortexes participate in network reconstruction when performing verbal fluency test. Only CFT differentiates the network reorganization between epilepsy groups and controls: augmented inter-hemispheric connections occur in R-FTE, whereas increased right intra-hemispheric efficacy happens in L-FTE. LFT parameters are better correlated with clinical conditions, especially in L-FTE. Funding: No funding