A Machine Learning Approach for the Identification of a Biomarker of Human Pain using fNIRS

Abstract

Pain is a highly unpleasant sensory and emotional experience, and no objective diagnosis test exists to assess it. In clinical practice there are two main methods for the estimation of pain, a patient’s self-report and clinical judgement. However, these methods are highly subjective and the need of biomarkers to measure pain is important to improve pain management, reduce risk factors, and contribute to a more objective, valid, and reliable diagnosis. Therefore, in this study we propose the use of functional near-infrared spectroscopy (fNIRS) and machine learning for the identification of a possible biomarker of pain. We collected pain information from 18 volunteers using the thermal test of the quantitative sensory testing (QST) protocol, according to temperature level (cold and hot) and pain intensity (low and high). Feature extraction was completed in three different domains (time, frequency, and wavelet), and a total of 69 features were obtained. Feature selection was carried out according to three criteria, information gain (IG), joint mutual information (JMI), and Chi-squared ( χ 2 ). The significance of each feature ranking was evaluated using three learning models separately, linear discriminant analysis (LDA), the K-nearest neighbour (K-NN) and support vector machines (SVM) using the linear and Gaussian and polynomial kernels. The results showed that the Gaussian SVM presented the highest accuracy (94.17%) using only 25 features to identify the four types of pain in our database. In addition, we propose the use of the top 13 features according to the JMI criteria, which exhibited an accuracy of 89.44%, as promising biomarker of pain. This study contributes to the idea of developing an objective assessment of pain and proposes a potential biomarker of human pain using fNIRS.

Electroencephalographic Workload Indicators During Teleoperation of an Unmanned Aerial Vehicle Shepherding a Swarm of Unmanned Ground Vehicles in Contested Environments

“…The selection criteria was based on the joint mutual information algorithm (JMI), this method ranks the features with the largest mutual information (MI) that produces most of the MI between the feature vector and the class label (Yang and Moody, 1999). The reason to choose JMI is that it presents better tradeoff in terms of accuracy, stability, and flexibility than other ranking methods (Brown et al, 2012;Rojas et al, 2019b). A disadvantage of this method is the fact that there is no stopping criteria to reach the best subset of features, and the user needs to select the number of features from the ranking list to form the optimal subset.…”

“…The objective of feature selection is to find a good representation of the data, improve estimators' performance by reducing the dimensionality of the data and eliminating redundant and irrelevant data from each participant's data (Rojas et al, 2019b). After applying joint mutual information (JMI), the features were ranked according to the their relevance to the class label.…”

Sensor Technologies to Manage the Physiological Traits of Chronic Pain: A Review

“…fNIRS is a non-invasive technique that has shown to be very useful in the evaluation of brain activity in response to pain stimuli. In [263], fNIRS was proposed as a biomarker of pain in conjunction with machine learning algorithms. Allodynia and motor loss of fibromyalgia patients has been related to alterations in the upper parietal gyrus bilateral region detected by fNIRS [264].…”

Detection and Comparative Analysis of Methylomic Biomarkers of Rheumatoid Arthritis

“…The chi-square test measures dependence between stochastic variables. It also checked whether a feature was statistically significantly associated with the class label under the assumption of a chi-squared distribution (Bangdiwala, 2016;Fernandez Rojas et al, 2019).…”