A research team led by Northwestern University faculty and alumni has found that it is possible to understand a patient’s pain level by analyzing data from vital signs.
In a new study, the artificial intelligence (AI) team, or machine learning, developed and applied algorithms to physiological data – including respiratory rate, blood pressure, heart rate, body temperature, and oxygen levels – from patients with persistent pain from sickle cell disease. Not only did the researchers’ approach improve the performance of baseline models to estimate subjective pain levels, it also found changes in pain and unusual pain variables.
The study will be published March 11 in the journal PLOS Computing Biology. This is the first paper that demonstrates that machine learning can be used to detect pain hidden within data from patient vital signs.
Currently, patients need to assess their pain based on a scale to zero to 10. This can be a daunting task as many people experience pain differently, and young children and unconscious patients cannot experience the pain. to consider between. The researchers believe that these thematic assessments of pain could be supplemented by a more focused, aggressive, data-driven approach to help physicians treat pain more accurately.
“The pain is subtle, so it’s hard to assess when trying to treat patients,” said Daniel Abrams of Northwestern, senior author of the study. “Doctors don’t want to give patients too much money and not provide enough pain relief. But they also don’t want to overdose on their patients because there is a risk of relapse. effects and additives. “
“Our study shows reasonable physiological data that is routinely collected at hospitals that are indicative of patient subjective pain,” said Mark Panaggio, the study’s first author. “We hope that our work will encourage the continued development of models for pain detection and ultimately prediction and that these models will allow clinicians to deliver faster and more focused treatments.”
Abrams is an associate professor of engineering sciences and applied mathematics at the McCormick School of Engineering at Northwestern. Panaggio was a former Ph.D. candidate from Abrams laboratory; he is now an applied mathematician at the Johns Hopkins University applied physics laboratory.
To conduct the study, the researchers used data from patients with sickle cell disease who were admitted to the hospital at Duke Medical Center because of debilitating pain. The sample included data from 105 hospitals in 46 special patients. When health care workers regularly collected patients ’vital signs, these patients also assessed their personal pain levels.
To simplify the operation, the researchers divided pain levels into three categories: low, medium and high. After using machine learning strategies to mine the data, the researchers compared their model assessment of pain with patients ’thematic reports.
“Our model findings reflected the thematic pain reports,” Abrams said. “It was even more accurate to determine if a patient was above or below their normal pain level.”
While hospital data can be difficult to obtain due to confidentiality issues, Abrams, Panaggio and their colleagues are currently receiving a much larger data set with pain reports from hundreds of thousands of patients. patients with pain due to cell wall disease and other causes, including postoperative pain and pain from unknown sources.
The researchers then aim to use their model to try to predict how pain relief might affect pain and predict when patients with chronic pain might experience it. discount, which is almost impossible to predict.
“A large proportion of people with chronic pain go to the emergency room with pain emergency events, in which pain is managed with prescription medications,” Abrams said. “Right now, no one knows what is causing these events. If we could predict these events, we could save patients a lot of pain and money.”
###
The study, “Can subjective pain be inferred from objective physical data? Evidence from patients with hedgehog disease,” was assisted by the National Institutes of Health (award number 1R01AT010413-01). In addition to Abrams and Panaggio, the paper’s coauthors include Fan Yang and Tanvi Banerjee of Wright State University and Nirmish Shah of Duke University.
Disclaimer: AAAS and EurekAlert! they are not responsible for the accuracy of press releases posted to EurekAlert! by sending institutions or for using any information through the EurekAlert system.