RESEARCH

HOME RESEARCH
Health Analytics
Clinical Attributes
Other: Computation Methods for Health
Speech and Language
Embedding stacked bottleneck vocal features in a LSTM architecture for automatic pain level classification during emergency triage
Abstract
In order to effectively allocate healthcare resource, a proper triage classification system plays an important role in assessing the severity of on-boarding patients at the emergency department. One of the major items in the current triage system is to assess the level of pain intensity, which relies solely on patients self-report numerical-rating scale (NRS) at the moment. The nature of self-report on pain level poses a challenge in maintaining the validity and consistency of the triage classification outcome. While there has been algorithms developed to automatically detect pain from expressive behaviors, most of them concentrate only on facial or body gestural expressions within the context of physical exercises. In this work, we propose to utilize stacked bottleneck acoustic representations in a long-short term memory neural networks (LSTMs) architecture as features for pain severity classification in a database consists of patients during real triage sessions. Our proposed framework achieves accuracy of 72.3% and 54.2% in binary and three-class pain intensity classification tasks. Our results further demonstrate that the severity of pain can largely be captured in the patients prosodic characteristics.
Figures
It shows the complete architecture of our stacked bottleneck vocal feature computation used for automatic pain classification: unsupervised learning with LSTM autoencoder to extract the bottleneck layer, fine-tuning the bottleneck layer on the patient’s acoustic data with supervised learning, and performing recognition with the fine-tuned outputted layer after functional encoding using support vector classification.
It shows the complete architecture of our stacked bottleneck vocal feature computation used for automatic pain classification: unsupervised learning with LSTM autoencoder to extract the bottleneck layer, fine-tuning the bottleneck layer on the patient’s acoustic data with supervised learning, and performing recognition with the fine-tuned outputted layer after functional encoding using support vector classification.
Authors
Chi-Chun Lee
Publication Date
2017/10/23
Conference
International Conference on Affective Computing and Intelligent Interaction (ACII)
2017 Seventh International Conference on Affective Computing and Intelligent Interaction (ACII)
DOI
10.1109/acii.2017.8273618
Publisher
IEEE