Title: Spontaneous back-pain alters randomness in functional connections in large scale brain networks: A random matrix perspective

Abstract: We use randomness as a measure to assess the impact of evoked pain on brain networks. Randomness is defined here as the intrinsic correlations that exist between different brain regions when the brain is in a task-free state. We use fMRI data of three brain states in a set of back pain patients monitored over a period of 6 months. We find that randomness in the task-free state closely follows the predictions of Gaussian orthogonal ensemble of random matrices. However, the randomness decreases when the brain is engaged in attending to painful inputs in patients suffering with early stages of back pain. A persistence of this pattern is observed in the patients that develop chronic back pain, while the patients who recover from pain after six months, the randomness no longer varies with the pain task. The study demonstrates the effectiveness of random matrix theory in differentiating between resting state and two distinct task states within the same patient. Further, it demonstrates that random matrix theory is effective in measuring systematic changes occurring in functional connectivity and offers new insights on how acute and chronic pain are processed in the brain at a network level.