Computer can detect pain, claims study


A new computer learning software can detect when people are in pain, claims a new US study.

A team at Stanford University in California have claimed that data gathered from brain scans can teach the computer through patterns of brain activity to determine whether or not someone was experiencing pain with 80% accuracy.

Doctors currently rely on patients communicating their symptoms, which can be unreliable, or in some cases impossible, regarding the very young, very old, or those who are unconscious.

Dr Sean Mackey, head of the study, said: “People have been looking for a pain detector for a very long time”.

For the study, eight volunteers underwent brain scans whilst touching hot objects, and hotter objects that were painful to touch.

The computer, which used linear support vector machinery – algorithms invented in 1995 – used the data to recognise the differences in patterns of brain activity.

“We’re hopeful we can eventually use this technology for better detection and better treatment of chronic pain.”