Classification of elderly pain severity from automated video clip facial action unit analysis: A study from a Thai data repository

Abstract

Data from 255 Thais with chronic pain were collected at Chiang Mai Medical School Hospital. After the patients self-rated their level of pain, a smartphone camera was used to capture faces for 10 s at a one-meter distance. For those unable to self-rate, a video recording was taken immediately after the move that causes the pain. The trained assistant rated each video clip for the pain assessment in advanced dementia (PAINAD). The pain was classified into three levels: mild, moderate, and severe. OpenFace© was used to convert the video clips into 18 facial action units (FAUs). Five classification models were used, including logistic regression, multilayer perception, naïve Bayes, decision tree, k-nearest neighbors (KNN), and support vector machine (SVM). Out of the models that only used FAU described in the literature (FAU 4, 6, 7, 9, 10, 25, 26, 27, and 45), multilayer perception is the most accurate, at 50%. The SVM model using FAU 1, 2, 4, 7, 9, 10, 12, 20, 25, and 45, and gender had the best accuracy of 58% among the machine learning selection features. Our open-source experiment for automatically analyzing video clips for FAUs is not robust for classifying pain in the elderly. The consensus method to transform facial recognition algorithm values comparable to the human ratings, and international good practice for reciprocal sharing of data may improve the accuracy and feasibility of the machine learning's facial pain rater.

Facial Analysis Technology for Pain Detection: A Potentially Useful Tool for People Living With Dementia

“…There are other AI technologies exploring automatic analysis of facial expression of pain, such as OpenFace, which in an open-source algorithm that looks at facial action units from the facial action coding system (FACS) in adults and children. 42,43 The OpenFace algorithm was trained using a Canadian pain dataset that is reported to be predominantly "young, healthy Caucasian persons�" 43 It may be at risk of some of the same challenges experienced by PainChek when used in different ethnic and cultural groups� For instance, OpenFace pain assessments were not correlated to PAINAD pain assessments in older Asian patients� 43 Researchers in Saskatchewan and Ontario have received a grant to develop technologies to help staff in long-term care homes monitor and record pain behaviour in older adults with dementia� 15 These technologies will detect facial expressions while residents are in their rooms, and alert staff if pain is detected in an individual� As part of this grant, the researchers have developed a fully automated computer model to detect pain level using facial expression� 44 This model was trained and evaluated using video data from 2 datasets, including both healthy, young adults and older adults with dementia� 44…”