An Intelligent Approach for Continuous Pain Intensity Prediction

We present a deep learning-based method for pain analysis using Long Short-Term Memory (LSTM) architecture to classify facial expressions of pain, demonstrating high accuracy in detecting various pain intensities, including subtle changes, with particular potential for clinical applications in assessing pain for non-verbal patients such as infants or those with cognitive impairments.

Conference ISBI

An Intelligent Approach for Continuous Pain Intensity Prediction

Facial expressions are an important indicator of pain, with different types of pain being associated with distinct facial expressions. Recent advances in deep learning have shown the potential for Machine learning techniques to automatically recognize and classify facial expressions of pain. In this study, we propose a method for pain analysis based on the LSTM architecture. The extracted features are fed into a LSTM network for pain level classification. To evaluate the proposed method, we conducted experiments on our comprehensive X-ITE dataset of facial expressions. The results showed that our method is capable of detecting different, even weak, pain intensities on the dataset with a high accuracy. Overall, our study demonstrates the potential of using the LSTM architecture for pain analysis based on facial expressions. The proposed method has important applications in clinical settings, such as pain assessment and management for patients who are unable to communicate their pain verbally, such as infants or individuals with cognitive impairments.

Fulltext Access

https://doi.org/10.1109/ISBI56570.2024.10635866

Citing

@inproceedings{OthmanMultimodalPain2022,
    title = {An Intelligent Approach for Continuous Pain Intensity Prediction},
    author = {Al-Radhi, Hassan and Fiedler, Marc-André and Al-Hamadi, Ayoub and Dinges, Laslo},
    booktitle = {2024 IEEE International Symposium on Biomedical Imaging (ISBI)},
    year = {2024}
    }