Enhanced Multimodal Physiological Signal Analysis for Pain Assessment Using Optimized Ensemble Deep Learning

e potential applications of multimodal physiological signals in healthcare, pain monitoring, and
clinical decision support systems have garnered signi cant attention in biomedical research. Subjective self-reporting
is the foundation of conventional pain assessment methods, which may be unreliable. Deep learning is a promising
alternative to resolve this limitation through automated pain classi cation. is paper proposes an ensemble deep
learning framework for pain assessment.
e framework makes use of features collected from electromyography
(EMG), skin conductance level (SCL), and electrocardiography (ECG) signals. We integrate Convolutional Neural
Networks (CNN), Long Short-Term Memory Networks (LSTM), Bidirectional Gated Recurrent Units (BiGRU), and
Deep Neural Networks (DNN) models. We then aggregate their predictions using a weighted averaging ensemble
technique to increase the classi cation’s robustness. To improve computing e ciency and remove redundant features,
we use Particle Swarm Optimization (PSO) for feature selection. is enables us to reduce the features’ dimensionality
without sacri cing the classi cation’s accuracy. With improved accuracy, precision, recall, and F-score across all pain
levels, the experimental results show that the suggested ensemble model performs better than individual deep learning
classi ers. In our experiments, the suggested model achieved over % accuracy, suggesting promising automated pain
assessment performance. However, due to di erences in validation protocols, comparisons with previous studies are
still limited. Combining deep learning and feature selection techniques signi cantly improves model generalization,
reducing over tting and enhancing classi cation performance. e evaluation was conducted using the BioVid Heat
Pain Dataset, con rming the model’s e ectiveness in distinguishing between di erent pain intensity levels.