Knowing how much pain a patient is in can help clinicians to diagnose a condition, understand the course of a disease, and set a course for treatment. While simply hearing patients out works pretty well for the general population, unconscious and non-communicative people are not properly assessed all too often.
Now, researchers from MIT, Harvard, and nearby hospitals have come up with what they believe is an objective method of measuring the pain that individuals experience using a non-invasive imaging technique.
They’re using functional near infrared spectroscopy to peer into the prefrontal cortex of the brain and assess the amount of activity going on within. Specifically, the technology measures the amount of oxygenated hemoglobin, which is indicative of the underlying neural activity.
Sensors are placed on the forehead to record the activity, but analyzing that data does not simply involve detecting higher activity levels, as pain generates specific activity patterns. To overcome this, the team used machine learning methods to identify the kind of brain activity that is indicative of pain, and used it as a biomarker of pain.
In their study, the team was able to identify that an individual was experiencing pain with 87% accuracy, which is already a remarkable accomplishment.
More research will be needed to confirm this technology in clinical practice and whether it is beneficial for unconscious, non-communicative, and even pediatric patients. Nevertheless, the potential for the objective analysis of pain as a significant clinical tool can’t be overstated.
Related paper in arXiv: Pain Detection with fNIRS-Measured Brain Signals