Researchers from one of South Korea's largest hospitals claim to have developed an objective measure of pain during and after surgery using AI.
FINDINGS
The research team at Asan Medical Center devised a pain assessment model that can be applied throughout the entire surgery process. It involves tracking a patient's heart rate, blood pressure, and blood volume change during surgery. A machine learning algorithm is utilised to analyse these parameters.
They demonstrated this model in a study involving 242 AMC surgery patients. Six features relating to pain prediction were selected and inputted into the AI-run model to confirm pain occurrence during and after surgery.
Findings, published in the NPJ Digital Medicine, showed that the AI-based model matched the accuracy of the existing model of intraoperative pain assessment at 83%. It, however, outperformed the latter at predicting postoperative pain at 93%, compared to its 58% accuracy.
Additionally, the study found the significance of new pain indicators, systolic upper limit variability and pulse width, which were previously not considered in the existing pain assessment model.
WHY IT MATTERS
Accurately assessing pain during and after surgery is crucial in patients' postoperative recovery.
The AMC researchers already developed a model for assessing intraoperative pain based on patient heart rate. However, an objective assessment remained wanting. According to them, pain intensity differs per patient – even for those undergoing the same operation, while unconscious patients or those under sedation are unable to express pain.
"This machine learning [model] will enable us to objectively evaluate the level of pain even in unconscious patients, such as those under sedation or those who have undergone endotracheal intubation," said Dr Byung-Moon Choi, professor at AMC Department of Anesthesiology and Pain Medicine and part of the research team.
"We expect this [AI-based model] to serve as important basic data for future personalised pain management."
THE LARGER TREND
Popular in the area of pain assessment,PainChek from Australia develops what could be the world's first mobile-based pain assessment application for seniors and children. Its technology also utilises AI and facial recognition to provide an objective measure of pain levels. The company is awaiting approval for its product from the United States Food and Drug Administration this year; the PainChek app is currently approved for use in Australia and Europe.
AI is also utilised in a new digital tool developed byNEC Corporation in Japan to estimate automatically the cause of chronic lower back pain.
Besides AI, virtual reality has also seen growing application in the area of pain management. US-basedAppliedVR offers a prescription VR-based device for treating chronic lower back pain. Another American company, XRHealth, has launched a VR game-based program to support patients in managing their pain.