Researchers at the ANESTHESIOLOGY 2023 annual meeting unveiled an AI-based Pain Recognition Tool, offering an unbiased method to assess pain levels before, during, and after surgery. Current subjective pain assessment tools may be influenced by biases, affecting patient care. This AI model, trained on facial images, demonstrated an 88% alignment with CPOT and a 66% alignment with VAS ratings. If validated, it could enhance pain management, potentially reducing anxiety, depression, and hospital stays. Privacy concerns will need addressing, with plans to incorporate additional monitoring features. AI holds promise in various surgical care applications, emphasizing the importance of patient safety.
An innovative AI Pain Recognition Tool has been developed to detect pain levels in patients both before, during, and after surgery, offering an impartial approach to assess pain and potentially reducing hospital stays.
At the ANESTHESIOLOGY 2023 annual meeting, researchers unveiled a groundbreaking automated pain recognition system powered by artificial intelligence (AI) that demonstrates the potential to identify pain in patients across various surgical phases.
Current methods of pain assessment often rely on subjective evaluations. The Critical-Care Pain Observation Tool (CPOT) necessitates healthcare teams to evaluate a patient’s pain based on indicators like body movement, muscle tension, and facial expressions. Meanwhile, the Visual Analog Scale (VAS) requires patients to self-report their pain, making these methods susceptible to biases and limited in their scope.
Timothy Heintz, BS, the lead author of the study and a fourth-year medical student at the University of California San Diego, explained, “Traditional pain assessment tools can be influenced by racial and cultural biases, potentially resulting in poor pain management and worse health outcomes. Furthermore, there is a gap in perioperative care due to the absence of continuous observable methods for pain detection. Our proof-of-concept AI model could help improve patient care through real-time, unbiased pain detection.”
Enhancing pain detection methods has the potential to prevent adverse outcomes such as anxiety and depression while also reducing the length of hospital stays, according to the researchers.
The AI model was trained on 143,293 facial images, comprising 159 non-pain episodes and 115 pain episodes in 69 patients who underwent various elective surgical procedures. The tool, which employs a combination of deep learning (DL) and computer vision, was trained to analyze each raw facial image and determine whether it represented pain or not.
After sufficient training, the model focused on specific facial expressions and muscles, particularly in areas like the lips, nose, and eyebrows.
Following training, the tool’s results aligned with VAS ratings 66 percent of the time and with CPOT ratings 88 percent of the time.
Heintz noted, “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. 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.”
If further studies validate the technology, the research team believes it could enhance pain assessment and management in clinical settings. In such a scenario, cameras placed on the ceiling and walls of a patient’s surgical recovery room could capture up to 15 images per second, which the AI tool would then analyze to assess pain levels in both conscious and unconscious patients.
Implementing this technology could also free up nurses and other members of the healthcare team from intermittent pain assessments, allowing them to focus on other aspects of patient care.
However, as this research progresses, privacy concerns must be addressed. The researchers intend to incorporate additional monitoring features, such as sound, movement, and brain and muscle activity, into the AI model.
This study represents just one of the potential applications of AI in surgical care. In a discussion with HealthITAnalytics in July, Desirée Chappell, CRNA, vice president of clinical quality at Northstar Anesthesia in Irving, Texas, and Jonathan Tan, MD, vice chair of Analytics and Clinical Effectiveness at the Children’s Hospital Los Angeles (CHLA), emphasized the significant promise of AI in anesthesiology. They also stressed the importance of navigating the hype surrounding these technologies while prioritizing patient safety.
