Multimodal signal fusion, most frequently combining electrodermal activity, electrocardiogram, and electromyography, provides a comprehensive approach to objectively assess acute pain.
Systematic Review
Integrating physiological signals from the heart, skin, and brain using deep learning offers significant opportunities to improve acute pain assessment systems.
Pain assessment poses unique challenges due to its subjective and multifaceted nature, often requiring the integration of various sensor modalities. This review aims to provide a comprehensive overview of recent research focused specifically on acute pain assessment, with specific attention to: (a) identifying combinations of sensor modalities utilised for pain assessment, (b) exploring methods for fusing data from diverse sensing modalities, and (c) examining the application of artificial intelligence (AI) methods for pain assessment using multimodal sensor data. A thorough literature search was conducted in September 2024, encompassing IEEE Xplore, Scopus, and Google Scholar databases, with a focus on articles published between January 2015 and September 2024. A total of 31 studies were included in this review, covering topics related to multimodal sensing, fusion techniques, and learning approaches. Notably, significant opportunities exist in integrating physiological signals, particularly from the heart, skin, and brain, by leveraging domain knowledge and deep learning methods to enhance the accuracy of pain monitoring systems. Furthermore, both the challenges and future directions for developing more effective pain assessment systems are discussed.
Khan et al. (Mon,) conducted a systematic review in Acute pain. Multimodal signal fusion was evaluated. Multimodal signal fusion, most frequently combining electrodermal activity, electrocardiogram, and electromyography, provides a comprehensive approach to objectively assess acute pain.