Machine Learning Unveils Clues to Chronic Pain in Brain Activity

Researchers at the University of California, San Francisco (UCSF), are making strides in the study and potential treatment of chronic pain through innovative methods. Dr. Prasad Shirvalkar and his team of doctors and researchers at the UCSF Pain Management Center have recently published a paper in Nature Neuroscience, detailing their investigation into how different types of chronic pain are represented in brain activity recordings.

Dr. Shirvalkar, a neurologist and pain specialist, focuses on central pain syndrome—a form of chronic pain caused by dysfunction in the central nervous system. Treating this condition is challenging because it does not originate from an external source or a damaged body part. Adding to the complexity is the limited understanding of how various types of pain, both acute and chronic, are represented in the nervous system. Dr. Shirvalkar explains that pain does not occur solely in specific brain regions but rather within distributed circuits of coordinated cells. While certain brain regions may contribute to specific types of information processing, they do not accomplish tasks independently.

One potential treatment for chronic neuropathic pain, such as central pain syndrome, is deep brain stimulation (DBS). DBS involves manipulating neural activity to alleviate painful sensations and is also employed for other neurological conditions like Parkinson's disease and severe depression. However, determining the optimal placement of DBS electrodes to effectively target the neural activity responsible for pain is challenging in chronic pain cases.

To gain insights into how central pain arises from brain function, the study focused on neural activity recorded from four patients with central pain syndrome. Three of the patients experienced chronic pain while recovering from a stroke, while the fourth experienced phantom limb pain after an above-the-knee amputation of the right leg. The researchers correlated the neural activity with questionnaires and pain scores reported by the patients on a scale from 0 to 10. Intracranial electrodes were implanted in the patients' orbitofrontal cortices (OFC) and anterior cingulate cortices (ACC), two brain regions previously associated with pain sensations.

The patients underwent surgery to have a portable neural recording and stimulation platform implanted in their brains. This device allowed them to continue with their daily activities while recording data representing their everyday experiences with central pain. The patients received reminders from a phone app to report their current pain level and record 30 seconds of brainwave data (electrocorticograms and local field potentials) from the neural implant.

Dr. Shirvalkar and his collaborators utilized machine learning techniques to establish connections between features of neural activity and the pain levels reported by each patient. These features, representing the relative presence of different frequencies in neural recordings from different locations in the ACC and OFC, consistently correlated with high and low pain states within each subject. However, the patterns of these associations—specific frequencies with the most power in different brain locations—varied significantly among patients.

Although the study observed differences in chronic pain-related brain activity between patients, the researchers identified commonalities. They found that in all patients, low-frequency power—specifically delta waves between 1 and 4 hertz—increased in lower OFC areas. These areas were located contralateral to the side of the body where the patients experienced pain. Dr. Shirvalkar emphasizes the surprising emergence of this contralateral OFC pattern, as all patients had pain restricted to one specific body part. He believes this presents an opportunity for further exploration, as brain laterality in relation to pain remains relatively unexplored.

Kate Nicholson, founder and executive director of the National Pain Advocacy Center, acknowledges the potential impact of this research in quantifying pain and improving treatment for individuals with central pain disorder and other chronic pain conditions. While recognizing the study's limited scope of only four participants and its focus on.

Reference:

https://spectrum.ieee.org/chronic-pain