We aim to study the neural mechanisms of inter-subject variability (i.e., people have different pain sensitivity) and intra-subject variability (i.e., pain perception fluctuates within an individual). In particular, we are interested in studying how the brain states (e.g., oscillations, BOLD responses) prior to pain stimuli would modulate forthcoming perception. We also aim to modulate brain oscillations using neuromodulation techniques to study the causality between the brain and behavior, and to move native observations to mechanistic manipulation.

Each chronic pain patient is marked by a unique contribution of sensory, emotional, cognitive, and motivational components to the experience of pain. Using multimodal neuroimaging data, advanced brain segmentation methods, and machine learning techniques, we develop core analytical approaches, including data fusion, feature selection, and connectomics analysis, to identify neural markers for chronic pain. We are also interested in building robust neuroimaging-based diagnostic and prognostic models for chronic pain, and develop models to predict treatment responses of neuromodulation approaches for chronic pain.

We use neuroimaging techniques to visualize the cerebral effects of different types of neuromodulation and then develop individualized protocols for chronic pain patients. To target different types of deficits (e.g. sensory, emotional, and cognitive) in chronic pain patients, we develop different types of brain and peripheral nerve stimulation.