FNIRS-Based Energy Landscape Analysis to Signify Brain Activity Dynamics of Individuals With Depression

Background: Major depressive disorder (MDD) is one of the most common mental disorders, and the number of individuals with MDD (MDDs) continues to increase. Therefore, there is an urgent need for an objective characterization and real-time detection method for depression. Functional near-infrared spectroscopy (fNIRS) is a non-invasive tool, which is widely used in depression research. However, the process of how the brain activity of MDDs changes in response to external stimuli based on fNIRS signals is not yet clear. Method: Energy landscape (EL) can describe the brain dynamics under task conditions by assigning energy values to each state. The higher the energy value, the lower the probability of the state occurring. This study compares the EL features of 60 MDDs with 60 healthy controls (HCs). Results: Compared to HCs, MDDs have more local minima, smaller energy differences, smaller variations in basin sizes, and longer duration in the basin of global minimum (GM). The classification results indicate that using the four features above for depression detection yields an accuracy of 86.53%. Simultaneously, there are significant differences between the two groups in the duration of the major states. Conclusion: The dynamic brain networks of MDDs exhibit more constraints and lower degrees of freedom, which might be associated with depressive symptoms such as negative emotional bias and rumination. In addition, we also demonstrate the strong depression detection capability of EL features, providing a possibility for their application in clinical diagnosis.