Altered brain dynamics in chronic neck and shoulder pain revealed by hidden Markov model

Chronic neck and shoulder pain (CNSP) is the most common clinical symptom of cervical spondylosis, which not only greatly affects individuals' quality of life but also places a significant burden on social healthcare systems. Existing analgesic treatments are often associated with significant adverse effects and limited efficacy. Recently, non-invasive neuromodulation techniques have shown promise, but the central mechanisms underlying chronic pain remain poorly understood. Recent advances in resting-state functional magnetic resonance imaging (rs-fMRI) have highlighted altered brain connectivity in CNSP patients. However, traditional methods, such as the sliding window approach, have limitations in capturing rapid fluctuations and individual differences in brain activity. The Hidden Markov Model (HMM) assumes that the brain is in different hidden states at different time points, with each state corresponding to a distinct connectivity pattern. It identifies state changes adaptively, without relying on preset time windows. In this study, we applied HMM to rs-fMRI data from CNSP patients and healthy controls to explore brain activity dynamics and state transition patterns. We identified five distinct brain states, revealing significant differences in functional occupancy, lifetime, switching rate, and state transition probabilities between CNSP patients and controls. This offers a novel neuroimaging perspective for personalizing interventions based on the individualized dynamic characteristics of CNSP patients. However, further research is needed to determine whether the number and nature of the internal states identified in this study can be generalized to other CNSP patient cohorts.