Toward a functional near-infrared spectroscopy-based monitoring of pain assessment for nonverbal patients

Abstract

Abstract. Pain diagnosis for nonverbal patients represents a challenge in clinical settings. Neuroimaging methods, such as functional magnetic resonance imaging and functional near-infrared spectroscopy (fNIRS), have shown promising results to assess neuronal function in response to nociception and pain. Recent studies suggest that neuroimaging in conjunction with machine learning models can be used to predict different cognitive tasks. The aim of this study is to expand previous studies by exploring the classification of fNIRS signals (oxyhaemoglobin) according to temperature level (cold and hot) and corresponding pain intensity (low and high) using machine learning models. Toward this aim, we used the quantitative sensory testing to determine pain threshold and pain tolerance to cold and heat in 18 healthy subjects (three females), mean age AE standard deviation (31.9 AE 5.5). The classification model is based on the bag-of-words approach, a histogram representation used in document classification based on the frequencies of extracted words and adapted for time series; two learning algorithms were used separately, K-nearest neighbor (K-NN) and support vector machines (SVM). A comparison between two sets of fNIRS channels was also made in the classification task, all 24 channels and 8 channels from the somatosensory region defined as our region of interest (RoI). The results showed that K-NN obtained slightly better results (92.08%) than SVM (91.25%) using the 24 channels; however, the performance slightly dropped using only channels from the RoI with K-NN (91.53%) and SVM (90.83%). These results indicate potential applications of fNIRS in the development of a physiologically based diagnosis of human pain that would benefit vulnerable patients who cannot self-report pain. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

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

“…There are two main techniques with appropriate temporal resolution to measure cognitive workload using brain signals: fNIRS and EEG. fNIRS measures cognitive workload by examining the levels of oxygenated (HbO) and deoxygenated (HbR) hemoglobin concentration in the cerebral cortex (Rojas et al, 2017b), and alertness, and indicative of loss of cortical arousal (Kamzanova et al, 2014). In this regard, fNIRS is commonly used to measure the amount of effort exerted in a given brain region in response to a given task.…”

Effects of Acupuncture Therapy on MCI Patients Using Functional Near-Infrared Spectroscopy

“…Functional near infrared spectroscopy (fNIRS) can be utilized as a non-invasive neuroimaging technique to monitor brain activity patterns by measuring concentration changes of oxygenated and deoxygenated hemoglobins (ΔHbO and ΔHbR) (Ferrari and Quaresima, 2012; Naseer and Hong, 2015). This neuroimaging method has been utilized to assess cortical activity changes in various research and clinical settings (Sakuma et al, 2014; Aasted et al, 2016; Rojas et al, 2017). Moreover, in comparison to well-established imaging methods such as fMRI, electroencephalograms (EEGs), and positron emission tomography (PET), fNIRS offers advantages such as portability, low cost, and lower susceptibility to movement artifacts.…”

Feasibility of a Real-Time Clinical Augmented Reality and Artificial Intelligence Framework for Pain Detection and Localization From the Brain

“…In this feasibility study, we assembled that pattern for data signature for objective clinical pain prediction using fNIRS data collected directly from the bilateral PFC and S1 cortices. Previous studies achieved promising results in attempting to classify different level of thermal stimulation (potentially indicating pain vs no-pain) based on hemodynamic response data collected from sensory cortex [25,26]. These studies also examined the performance of several prevalent machine learning methods including support vector machine, linear discriminant analysis, and K-nearest neighbor.…”