Enhanced residual attention-based subject-specific network (ErAS-Net): facial expression-based pain classification with multiple attention mechanisms

• Published in: Scientific reports Vol. 15; no. 1; pp. 19425 - 16
• Main Authors: Morsali, Mahdi, Ghaffari, Aboozar
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 03.06.2025; Nature Publishing Group; Nature Portfolio
• Subjects: 639/166/985; 639/166/987; Accuracy; Algorithms; Attention; Attention mechanisms; Classification; Deep Learning; Facial deformation; Facial Expression; Humanities and Social Sciences; Humans; multidisciplinary; Pain; Pain - classification; Pain - diagnosis; Pain detection; Pain Measurement - methods; Pain perception; Science; Science (multidisciplinary); Transfer learning; Deep learning; Attention mechanisms; Facial deformation; Transfer learning; Facial expression; Pain detection
• DOI: 10.1038/s41598-025-04552-w
• ISSN: 2045-2322

The automatic detection of pain through the analysis of facial expressions is indeed one of the most critical challenges in the healthcare system. One of the significant challenges in automatic pain detection from facial expressions is the variability in how individuals express pain and other emotions through their facial deformations. This research aims to solve this issue by presenting ErAS-Net, an Enhanced Residual Attention-Based Subject-Specific Network that employs various attention mechanisms. Through transfer learning and multiple attention mechanisms, the proposed deep learning model is designed to mimic human perception of facial expressions, thereby enhancing its pain recognition ability and capturing the unique features of each individual’s facial expressions based on their specific patterns. The UNBC-McMaster Shoulder Pain dataset is used to demonstrate the effectiveness of the proposed deep learning algorithm, which achieves impressive values of 98.77% accuracy for binary classification and 94.21% for four-level pain intensity classification using tenfold cross-validation. Additionally, the model attained 89.83% accuracy for binary classification with the Leave-One-Subject-Out (LOSO) validation method. To further evaluate generalizability, a cross-dataset experiment was conducted using the BioVid Heat Pain Database, where ErAS-Net achieved 78.14% accuracy for binary pain detection on unseen data without fine-tuning. The fact that this finding supports the attention mechanism and human perception is why the proposed model proves to be a powerful and reliable tool for automatic pain detection.