Efficiency of facial expression diagnostic system in pain assessment after geriatric surgery

With an increasingly ageing population, it is critical to understand how the systems in our body change with age and how this affects pain assessment. The aim of this methodological study was to evaluate the effectiveness of the facial expression recognition system in geriatric postoperative pain assessment. The study was conducted between October 2022 and October 2024 in the general surgery service of Zonguldak Atatürk State Hospital. The sample of the study consisted of 68 patients aged 65 years and over in the postoperative period. Data were collected with the "Sociodemographic Data Collection Form", "Numerical Rating Scale" and "Wong Baker Faces Pain Scale". Immediately after the patient left the operating room and came to the inpatient ward, two follow-ups were made by recording 30 seconds with a video camera, at 0 minute and 1 hour. At the same time, the patient and the nurse filled out the forms. Video recording, pain intensity scores given by the patient and the nurse were recorded on the computer. Facial expressions of the patients were coded with the Facial Action Coding System. Pain intensity scores given by the patient in MatLab were combined with facial action units. According to the scores given by the patient and the artificial neural network in both measurements, it was determined that there was a statistically significant relationship between the patient and the artificial neural network and that they showed a very high level of agreement in categorical classification (κ>0.81, p<0.001). The facial expression recognition system successfully performed pain assessment after geriatric surgery. It is recommended that current technologies should be used in pain assessment, artificial intelligence should be utilised and nurses should take a more active role.