Chapter 10 - EEG biomarkers of pain and applications of machine learning

• Aleksandra Vuckovic, Mohammed Sabah Jarjees, Muhammad Abul Hasan, Mariel Purcell, Matthew Fraser

Sensory mechanical and thermal tests are widely used biomarkers of Central Neuropathic Pain (CNP) in people with Spinal Cord Injury. They are used to diagnose, and in some cases to predict the risk of developing CNP. In this chapter, we propose neuroimaging EEG biomarkers for diagnosis and prediction of CNP in people with chronic and subacute CNP. EEG biomarkers utilize inexpensive technology and do not rely on the level of preservation of sensation. Due to its ability to record dynamically changing events on the order of milliseconds, EEG is suitable for creating biomarkers in relaxed state as well as during a cognitive function, which can complement each other. We provide a review of the state-of-the-art EEG markers of CNP followed by several examples from our own research. We present diagnostic and prognostic EEG biomarkers of CNP based on machine learning, derived from EEG in resting state and during imagined movements of paralyzed and nonparalyzed limbs in people with subacute and chronic SCI.