Physiological Signal-Based Method for Measurement of Pain Intensity

Abstract

The standard method for prediction of the absence and presence of pain has long been self-report. However, for patients with major cognitive or communicative impairments, it would be better if clinicians could quantify pain without having to rely on the patient's self-description. Here, we present a newly pain intensity measurement method based on multiple physiological signals, including blood volume pulse (BVP), electrocardiogram (ECG), and skin conductance level (SCL), all of which are induced by external electrical stimulation. The proposed pain prediction system consists of signal acquisition and preprocessing, feature extraction, feature selection and feature reduction, and three types of pattern classifiers. Feature extraction phase is devised to extract pain-related characteristics from short-segment signals. A hybrid procedure of genetic algorithm-based feature selection and principal component analysis-based feature reduction was established to obtain high-quality features combination with significant discriminatory information. Three types of classification algorithms-linear discriminant analysis, k-nearest neighbor algorithm, and support vector machine-are adopted during various scenarios, including multi-signal scenario, multi-subject and between-subject scenario, and multi-day scenario. The classifiers gave correct classification ratios much higher than chance probability, with the overall average accuracy of 75% above for four pain intensity. Our experimental results demonstrate that the proposed method can provide an objective and quantitative evaluation of pain intensity. The method might be used to develop a wearable device that is suitable for daily use in clinical settings.

Keywords: feature extraction; feature selection and reduction; pain intensity quantification; pattern classification; physiological signals.

Figure 1

(A) Pain recording scenario. Positioning of sensors and electrical stimulation electrodes: (1) BVP-Flex/Pro sensor, (2) EKG-Flex/Pro sensor, (3) SC-Flex/Pro sensor, and (4) Electrical stimulation electrodes. (B) Pain induction protocol. The stimulus levels are represented by current intensity L1 to L3. The features are extracted from the green window of length 30 s.

Figure 2

Physiological signals of a subject at baseline, stim20, stim30, and stim40. From top to bottom: blood volume pressure (BVP; percent reflectance), electrocardiogram (ECG; microVoltage), and skin conductance (SCL; microSiemens). Each graph shows 10 s of response. The segments shown here are visibly different for the four states.

Figure 3

Processing pipeline of the pain intensity recognition system.

Figure 4

The classification accuracy of the LDA algorithm for each physiological signal in each of the six subjects.

Figure 5

The classification accuracy of running ten times LDA for all 36 features in each subject.

Figure 6

An example sample set from subject 1 before dimension reduction projected onto the first three features.

Figure 7

An example sample set from subject 1 after dimension reduction projected onto the first three features.

Figure 8

The classification accuracy of three classifiers (KNN, SVM and LDA) after feature processing in each of the six subjects.

Figure 9

The classification accuracy of running ten times three classifiers (KNN, SVM and LDA) for a multiple-subject scenario.

Figure 10

Pain intensity recognition of three classifiers (KNN, SVM and LDA) for a between-subject scenario.

Figure 11

Pain intensity recognition of three classifiers (KNN, SVM and LDA) for a multi-day scenario.