Intrinsic neural timescales during mental tasks: An EEG test-retest reliability study

Intrinsic neural timescales (INT) describe the storage time of neural information in local brain regions, representing the time window for information integration and segregation. Mental task-state electroencephalography (EEG) holds potential clinical value for patients; however, the test-retest reliability of INT remains an open issue. The present study aimed to explore the spatial characteristics of INT and investigate its test-retest reliability by analyzing the data from three mental tasks-subtraction, memory, and music-derived from an open-source EEG database. A classic spatial distribution hierarchy was demonstrated. Our results also revealed that INT had moderate-to-good test-retest reliability. The test-retest reliability of INT in the subtraction task was higher than that in memory and music tasks. Autocorrelation window (ACW)-0 exhibited superior test-retest reliability compared to ACW-50. In summary, our findings provide robust validation of the test-retest reliability of the mental task-state EEG INT from multiple perspectives, thereby facilitating its application in both basic research and clinical settings.