Topics:

Stanford algorithm eases ache of diagnosing chronic pain

Tools

How do we really know a patient's level of pain? Today, doctors ask patients to rate their own pain on a scale, relying heavily on what patients say to make their diagnoses. Stanford researchers are working on a more objective measurement, and they have applied computer algorithms to help them pull it off.

The researchers used advanced algorithms to analyze structural data in brain scans, Stanford reported on Monday. In a recent study, the algorithms accurately predicted whether patients suffered from lower back pain 76% of the time. And this has inspired the researchers to seek improvements on the method in an effort to provide doctors with more than self-reported data from patients to diagnose chronic pain.

An objective method of diagnosing chronic pain could be a boon in healthcare, especially when treating patients such as young children and elders who are unable to tell clinicians about their pain. And as the San Francisco Business Times reports, such a system could also benefit drug development. However, the long-standing hunt for an objective measure of pain appears to have no immediate end in sight.

 "We're still a long way from that," Stanford's Dr. Sean Mackey, chief of the Division of Pain Management, stated, "but this method may someday augment self-reporting as the primary way of determining whether a patient is in chronic pain."

Mackey and his colleagues from Stanford University School of Medicine proposed the use of pattern-recognition algorithms to analyze magnetic resonance imaging (MRI) scans from both healthy and ill patients to arrive at a model of pain. Mackey indicated to the Business Times that he's interested in how data from genetic tests and other sources could be used to refine ways to make an objective diagnosis.

- see the item from Stanford
- and the Business Times article

Related Articles:
Novartis predicts drug side effects with computers
Stanford algorithm uncovers drug side effects
MRI technique may serve as chronic pain biomarker