An automated system for pain recognition using artificial intelligence (AI) is appearing effective as an impartial method for detecting pain in patients before, during, and after surgery, as revealed in research presented at the ANESTHESIOLOGY® 2023 annual meeting.
Nairametrics learns that currently, subjective methods like the Visual Analog Scale (VAS) and the Critical-Care Pain Observation Tool (CPOT) are employed to assess pain.
However, with this automated pain recognition system, two AI techniques, computer vision and deep learning, are incorporated allowing it to interpret visual cues to assess patients’ pain.
Early recognition and effective management of pain have been demonstrated to reduce hospital stays and prevent long-term health conditions like chronic pain, anxiety, and depression.
How the AI system detects pain
In the study, researchers provided the AI model with 143,293 facial images from 115 pain episodes and 159 non-pain episodes in 69 patients undergoing a variety of elective surgical procedures.
The AI learned to identify patterns by analyzing raw facial images, with a focus on facial expressions and muscles in specific facial areas like the eyebrows, lips, and nose.
With enough examples, it made accurate pain predictions. The AI-automated pain recognition system aligned with CPOT results 88% of the time and with VAS results 66% of the time.
- “The VAS is less accurate compared to CPOT because VAS is a subjective measurement that can be more heavily influenced by emotions and behaviors than CPOT might be,” said Heintz, one of the researchers.
- “However, our models were able to predict VAS to some extent, indicating there are very subtle cues that the AI system can identify that humans cannot.”
What next…
If the findings are confirmed, this technology could serve as an additional tool for physicians to enhance patient care.
For instance, cameras could be strategically placed on the walls and ceilings of the post-anesthesia care unit to continuously assess patients’ pain, including those who are unconscious, capturing 15 images per second.
This approach could also alleviate the workload of nurses and healthcare professionals who periodically evaluate patients’ pain, allowing them to concentrate on other aspects of care.
The research team intends to further expand the model by incorporating additional variables like patient movement and sound.
Addressing privacy concerns would be crucial to safeguard patient images. In the future, the system could potentially include other monitoring features, such as assessing the brain and muscle activity of unconscious patients.
I wonder if this AI pain detector system can be used in other specialized areas, like psychology.