Comparison of Feature Extraction Methods for Physiological Signals for Heat-Based Pain Recognition

Autonomic responses to pressure sensitivity of head, face and neck: Heart rate and skin conductance

Subjective scales are frequently used in the design process of head-related products to assess pressure discomfort. Nevertheless, some users lack fundamental cognitive and motor abilities (e.g., paralyzed patients). Therefore, it is vital to find non-verbal measurements of pressure discomfort and pressure pain. This study gathered the autonomic response data (heart rate and skin conductance) of 30 landmarks in head, neck and face from 31 participants experiencing pressure discomfort and pressure pain. The results indicate that pressure stimulation can change heart rate (HR) and skin conductance (SC). SC can be more useful in assessing pressure discomfort than HR for specific landmarks, and SC also possesses a faster arousal rate than HR. Moreover, HR decreased in response to pressure stimulation, while SC decreased followed by an increase. In comparisons between genders, the subjective pressure discomfort threshold (PDT) and pressure pain threshold (PPT) of women were lower than those of men, but men's autonomic responses (HR and SC) were more intense. Furthermore, there was no linear correlation between subjective pressure thresholds (PDT and PPT) and autonomic response intensity. This study has significant implications for resolving ergonomic issues (pressure discomfort and pain) associated with head-related products.

Transformer encoder with multiscale deep learning for pain classification using physiological signals

Pain, a pervasive global health concern, affects a large segment of population worldwide. Accurate pain assessment remains a challenge due to the limitations of conventional self-report scales, which often yield inconsistent results and are susceptible to bias. Recognizing this gap, our study introduces PainAttnNet, a novel deep-learning model designed for precise pain intensity classification using physiological signals. We investigate whether PainAttnNet would outperform existing models in capturing temporal dependencies. The model integrates multiscale convolutional networks, squeeze-and-excitation residual networks, and a transformer encoder block. This integration is pivotal for extracting robust features across multiple time windows, emphasizing feature interdependencies, and enhancing temporal dependency analysis. Evaluation of PainAttnNet on the BioVid heat pain dataset confirm the model's superior performance over the existing models. The results establish PainAttnNet as a promising tool for automating and refining pain assessments. Our research not only introduces a novel computational approach but also sets the stage for more individualized and accurate pain assessment and management in the future.

Objective Measurement of Subjective Pain Perception with Autonomic Body Reactions in Healthy Subjects and Chronic Back Pain Patients: An Experimental Heat Pain Study

Multiple attempts to quantify pain objectively using single measures of physiological body responses have been performed in the past, but the variability across participants reduces the usefulness of such methods. Therefore, this study aims to evaluate whether combining multiple autonomic parameters is more appropriate to quantify the perceived pain intensity of healthy subjects (HSs) and chronic back pain patients (CBPPs) during experimental heat pain stimulation. HS and CBPP received different heat pain stimuli adjusted for individual pain tolerance via a CE-certified thermode. Different sensors measured physiological responses. Machine learning models were trained to evaluate performance in distinguishing pain levels and identify key sensors and features for the classification task. The results show that distinguishing between no and severe pain is significantly easier than discriminating lower pain levels. Electrodermal activity is the best marker for distinguishing between low and high pain levels. However, recursive feature elimination showed that an optimal subset of features for all modalities includes characteristics retrieved from several modalities. Moreover, the study's findings indicate that differences in physiological responses to pain in HS and CBPP remain small.

Design and Evaluation of Deep Learning Models for Continuous Acute Pain Detection Based on Phasic Electrodermal Activity

The current method for assessing pain in clinical practice is subjective and relies on self-reported scales. An objective and accurate method of pain assessment is needed for physicians to prescribe the proper medication dosage, which could reduce addiction to opioids. Hence, many works have used electrodermal activity (EDA) as a suitable signal for detecting pain. Previous studies have used machine learning and deep learning to detect pain responses, but none have used a sequence-to-sequence deep learning approach to continuously detect acute pain from EDA signals, as well as accurate detection of pain onset. In this study, we evaluated deep learning models including 1-dimensional convolutional neural networks (1D-CNN), long short-term memory networks (LSTM), and three hybrid CNN-LSTM architectures for continuous pain detection using phasic EDA features. We used a database consisting of 36 healthy volunteers who underwent pain stimuli induced by a thermal grill. We extracted the phasic component, phasic drivers, and time-frequency spectrum of the phasic EDA (TFS-phEDA), which was found to be the most discerning physiomarker. The best model was a parallel hybrid architecture of a temporal convolutional neural network and a stacked bi-directional and uni-directional LSTM, which obtained a F1-score of 77.8% and was able to correctly detect pain in 15-second signals. The model was evaluated using 37 independent subjects from the BioVid Heat Pain Database and outperformed other approaches in recognizing higher pain levels compared to baseline with an accuracy of 91.5%. The results show the feasibility of continuous pain detection using deep learning and EDA.

Automatic assessment of pain based on deep learning methods: A systematic review

BACKGROUND AND OBJECTIVE

The automatic assessment of pain is vital in designing optimal pain management interventions focused on reducing suffering and preventing the functional decline of patients. In recent years, there has been a surge in the adoption of deep learning algorithms by researchers attempting to encode the multidimensional nature of pain into meaningful features. This systematic review aims to discuss the models, the methods, and the types of data employed in establishing the foundation of a deep learning-based automatic pain assessment system.

METHODS

The systematic review was conducted by identifying original studies searching digital libraries, namely Scopus, IEEE Xplore, and ACM Digital Library. Inclusion and exclusion criteria were applied to retrieve and select those of interest, published until December 2021.

RESULTS

A total of one hundred and ten publications were identified and categorized by the number of information channels used (unimodal versus multimodal approaches) and whether the temporal dimension was also used.

CONCLUSIONS

This review demonstrates the importance of multimodal approaches for automatic pain estimation, especially in clinical settings, and also reveals that significant improvements are observed when the temporal exploitation of modalities is included. It provides suggestions regarding better-performing deep architectures and learning methods. Also, it provides suggestions for adopting robust evaluation protocols and interpretation methods to provide objective and comprehensible results. Furthermore, the review presents the limitations of the available pain databases for optimally supporting deep learning model development, validation, and application as decision-support tools in real-life scenarios.

Explainable Artificial Intelligence (XAI) in Pain Research: Understanding the Role of Electrodermal Activity for Automated Pain Recognition

Artificial intelligence and especially deep learning methods have achieved outstanding results for various applications in the past few years. Pain recognition is one of them, as various models have been proposed to replace the previous gold standard with an automated and objective assessment. While the accuracy of such models could be increased incrementally, the understandability and transparency of these systems have not been the main focus of the research community thus far. Thus, in this work, several outcomes and insights of explainable artificial intelligence applied to the electrodermal activity sensor data of the PainMonit and BioVid Heat Pain Database are presented. For this purpose, the importance of hand-crafted features is evaluated using recursive feature elimination based on impurity scores in Random Forest (RF) models. Additionally, Gradient-weighted class activation mapping is applied to highlight the most impactful features learned by deep learning models. Our studies highlight the following insights: (1) Very simple hand-crafted features can yield comparative performances to deep learning models for pain recognition, especially when properly selected with recursive feature elimination. Thus, the use of complex neural networks should be questioned in pain recognition, especially considering their computational costs; and (2) both traditional feature engineering and deep feature learning approaches rely on simple characteristics of the input time-series data to make their decision in the context of automated pain recognition.

Using the Electrocardiogram for Pain Classification under Emotional Contexts

The adequate characterization of pain is critical in diagnosis and therapy selection, and currently is subjectively assessed by patient communication and self-evaluation. Thus, pain recognition and assessment have been a target of study in past years due to the importance of objective measurement. The goal of this work is the analysis of the electrocardiogram (ECG) under emotional contexts and reasoning on the physiological classification of pain under neutral and fear conditions. Using data from both contexts for pain classification, a balanced accuracy of up to 97.4% was obtained. Using an emotionally independent approach and using data from one emotional context to learn pain and data from the other to evaluate the models, a balanced accuracy of up to 97.7% was reached. These similar results seem to support that the physiological response to pain was maintained despite the different emotional contexts. Attempting a participant-independent approach for pain classification and using a leave-one-out cross-validation strategy, data from the fear context were used to train pain classification models, and data from the neutral context were used to evaluate the performance, achieving a balanced accuracy of up to 94.9%. Moreover, across the different learning strategies, Random Forest outperformed the remaining models. These results show the feasibility of identifying pain through physiological characteristics of the ECG response despite the presence of autonomic nervous system perturbations.

Analysis of Physiological Responses during Pain Induction

Pain is a complex phenomenon that arises from the interaction of multiple neuroanatomic and neurochemical systems with several cognitive and affective processes. Nowadays, the assessment of pain intensity still relies on the use of self-reports. However, recent research has shown a connection between the perception of pain and exacerbated stress response in the Autonomic Nervous System. As a result, there has been an increasing analysis of the use of autonomic reactivity with the objective to assess pain. In the present study, the methods include pre-processing, feature extraction, and feature analysis. For the purpose of understanding and characterizing physiological responses of pain, different physiological signals were, simultaneously, recorded while a pain-inducing protocol was performed. The obtained results, for the electrocardiogram (ECG), showed a statistically significant increase in the heart rate, during the painful period compared to non-painful periods. Additionally, heart rate variability features demonstrated a decrease in the Parasympathetic Nervous System influence. The features from the electromyogram (EMG) showed an increase in power and contraction force of the muscle during the pain induction task. Lastly, the electrodermal activity (EDA) showed an adjustment of the sudomotor activity, implying an increase in the Sympathetic Nervous System activity during the experience of pain.

Temporal properties of pain contrast enhancement using repetitive stimulation

BACKGROUND

Offset analgesia (OA) is characterized by a disproportionately large reduction in pain following a small decrease in noxious stimulation and is based on temporal pain contrast enhancement (TPCE). The underlying mechanisms of this phenomenon are still poorly understood. This study is aiming to investigate whether TPCE can also be induced by repetitive stimulation, i.e., by stimuli clearly separated in time.

METHODS

A repetitive TPCE paradigm was induced in healthy, pain-free subjects (n = 33) using heat stimuli. Three different interstimulus intervals (ISIs) were used: 5, 15, and 25 seconds. All paradigms were contrasted with a control paradigm without temperature change. Participants continuously rated perceived pain intensity. In addition, electrodermal activity (EDA) was recorded as a surrogate measure of autonomic arousal.

RESULTS

Temporal pain contrast enhancement was confirmed for ISI 5 seconds (p < 0.001) and ISI 15 seconds (p = 0.005) but not for ISI 25 seconds (p = 0.07), however, the magnitude of TPCE did not differ between ISIs (p = 0.11). A TPCE-like effect was also detected with increased EDA values.

CONCLUSIONS

TPCE can be induced by repetitive stimulation. This finding may be explained by a combination of the mechanisms underlying the OA and a facilitated pain habituation.

A Machine Learning Framework for Automated Accident Detection Based on Multimodal Sensors in Cars

Identifying accident patterns is one of the most vital research foci of driving analysis. Environmental or safety applications and the growing area of fleet management all benefit from accident detection contributions by minimizing the risk vehicles and drivers are subject to, improving their service and reducing overhead costs. Some solutions have been proposed in the past literature for automated accident detection that are mainly based on traffic data or external sensors. However, traffic data can be difficult to access, while external sensors can end up being difficult to set up and unreliable, depending on how they are used. Additionally, the scarcity of accident detection data has limited the type of approaches used in the past, leaving in particular, machine learning (ML) relatively unexplored. Thus, in this paper, we propose a ML framework for automated car accident detection based on mutimodal in-car sensors. Our work is a unique and innovative study on detecting real-world driving accidents by applying state-of-the-art feature extraction methods using basic sensors in cars. In total, five different feature extraction approaches, including techniques based on feature engineering and feature learning with deep learning are evaluated on the strategic highway research program (SHRP2) naturalistic driving study (NDS) crash data set. The main observations of this study are as follows: (1) CNN features with a SVM classifier obtain very promising results, outperforming all other tested approaches. (2) Feature engineering and feature learning approaches were finding different best performing features. Therefore, our fusion experiment indicates that these two feature sets can be efficiently combined. (3) Unsupervised feature extraction remarkably achieves a notable performance score.

Physiological responses to pain in cancer patients: A systematic review

BACKGROUND AND OBJECTIVE

Pain is one of the most debilitating symptoms in persons with cancer. Still, its assessment is often neglected both by patients and healthcare professionals. There is increasing interest in conducting pain assessment and monitoring via physiological signals that promise to overcome the limitations of state-of-the-art pain assessment tools. This systematic review aims to evaluate existing experimental studies to identify the most promising methods and results for objectively quantifying cancer patients' pain experience.

METHODS

Four electronic databases (Pubmed, Compendex, Scopus, Web of Science) were systematically searched for articles published up to October 2020.

RESULTS

Fourteen studies (528 participants) were included in the review. The selected studies analyzed seven physiological signals. Blood pressure and ECG were the most used signals. Sixteen physiological parameters showed significant changes in association with pain. The studies were fairly consistent in stating that heart rate, the low-frequency to high-frequency component ratio (LF/HF), and systolic blood pressure positively correlate with the pain.

CONCLUSIONS

Current evidence supports the hypothesis that physiological signals can help objectively quantify, at least in part, cancer patients' pain experience. While there is much more to be done to obtain a reliable pain assessment method, this review takes an essential first step by highlighting issues that should be taken into account in future research: use of a wearable device for pervasive recording in a real-world context, implementation of a big-data approach possibly supported by AI, including multiple stratification factors (e.g., cancer site and stage, source of pain, demographic and psychosocial data), and better-defined recording procedures. Improved methods and algorithms could then become valuable add-ons in taking charge of cancer patients.