Measurement of Pain Response in Geriatric Dentistry using the COMFORT Behavioral Algorithm
Keywords:
Geriatric dentistry, Pain Assessment, Comfort Behavioral Algorithm, Artificial Intelligence, Facial Analysis, Logistic Regression, Non-Verbal Patients.Abstract
Geriatric dentistry is strongly encouraged to assess pain accurately in patients with compromised
cognition, difficulties in communication, and sensory changes commonly associated with the elderly.
Based on this, this paper proposes an AI system that utilizes the COMFORT Behavioral Algorithm,
initially developed for assessing pain in non-verbal pediatric patients, to quantify pain in elderly patients
undergoing dental treatment. A real-time observational system tracked facial, postural, and vocal cues,
processing them to identify behavioral traits such as alertness, calmness, muscle tone, and facial tension.
They were weighed and normalized into a COMFORT Score, and a logistic regression model predicted
the probability of pain occurrence. Clinician feedback in real-time and automatic pain scoring were
enabled through the system architecture. Results in 60 patients indicated that facial tension and muscle
tone were the most predictive indicators of pain, and the model was found to be highly consistent with
human expert ratings, with an RMSE of 0.41. This research supports the viability of AI-facilitated
behavioral scoring to enhance evaluation and treatment of pain in non-verbal elderly populations.
