fNIRS-based brain-computer interfaces: a review.

Naseer N, Hong KS. · 2015

Read fNIRS-based brain-computer interfaces: a review. by Naseer N, Hong KS. (2015) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4309034.txt.

Introduction

Noman Naseer and Keum-Shik Hong review brain-computer interfaces built on near-infrared light scans of the head. Their review appeared in Frontiers in Human Neuroscience in 2015. Such interfaces let people control devices with brain activity alone, bypassing muscles, which matters most for people with severe motor disability.

The method, fNIRS, tracks blood flow changes through the skull with light. The review covers the mental tasks users perform, the brain areas most often read, and the full processing chain from cleaning noise to classifying intent. Motor imagery from the motor cortex and mental arithmetic from the front of the brain are the workhorse tasks. Strengths include low cost and comfort; speed and accuracy still trail implanted systems.

Catalog record

CREATORNaseer N, Hong KS.
YEAR2015
COLLECTIONPAPERS
FORMATTXT
FILE SIZE73 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC4309034

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

fNIRS-based brain-computer interfaces: a review By Noman Naseer, Keum-Shik Hong Published in: Frontiers in Human Neuroscience DOI: 10.3389/fnhum.2015.00003 ABSTRACT A brain-computer interface (BCI) is a communication system that allows the use of brain activity to control computers or other external devices. It can, by bypassing the peripheral nervous system, provide a means of communication for people suffering from severe motor disabilities or in a persistent vegetative state. In this paper, brain-signal generation tasks, noise removal methods, feature extraction/selection schemes, and classification techniques for fNIRS-based BCI are reviewed. The most common brain areas for fNIRS BCI are the primary motor cortex and the prefrontal cortex. In relation to the motor cortex, motor imagery tasks were preferred to motor execution tasks since possible proprioceptive feedback could be avoided. In relation to the prefrontal cortex, fNIRS showed a significant advantage due to no hair in detecting the cognitive tasks like mental arithmetic, music imagery, emotion induction, etc. In removing physiological noise in fNIRS data, band-pass filtering was mostly used. However, more advanced techniques like adaptive filtering, independent component analysis (ICA), multi optodes arrangement, etc. are being pursued to overcome the problem that a band-pass filter cannot be used when both brain and physiological signals occur within a close band. In extracting features related to the desired brain signal, the mean, variance, peak value, slope, skewness, and kurtosis of the noised-removed hemodynamic response were used. For classification, the linear discriminant analysis method provided simple but good performance among others: support vector machine (SVM), hidden Markov model (HMM), artificial neural network, etc. fNIRS will be more widely used to monitor the occurrence of neuro-plasticity after neuro-rehabilitation and neuro-stimulation. Technical breakthroughs in the future are expected via bundled-type probes, hybrid EEG-fNIRS BCI, and through the detection of initial dips. INTRODUCTION A brain-computer interface (BCI) system provides its users with control channels that are independent of the brain's output channels (i.e., the peripheral nervous system and muscles) (Wolpaw et al., 2002). Such systems can be used as a means for communications and restoration of motor functions (through a neuroprosthesis) for people with motor disorders such as amyotrophic lateral sclerosis (ALS) and spinal cord injury, and/or people in the persistent locked-in state (LIS). It can also be used as a neurorehabilitation tool to improve motor and/or cognitive performance of such people. A typical BCI system consists of five stages (see Figure 1): brain-signal acquisition, preprocessing, feature extraction/selection, classification, and application interface. In the first brain-signal acquisition stage, suitable signals are acquired using an appropriate brain-imaging modality. Since the acquired signals are normally weak and contain noises (physiological and instrumental) and artifacts, preprocessing is needed, which is the second stage. In the third stage, some useful data so called “features” are extracted. These features, in the fourth stage, are classified using a suitable classifier. Finally, in the fifth stage, the classified signals are transmitted to a computer or other external devices for generating the desired control commands to the devices. In neurofeedback applications, a real-time display of brain activity is desirable, which enables self-regulation of brain functions. Figure 1 depicts a schematic of (hybrid) functional near-infrared spectroscopy (fNIRS) and electroencephalography (EEG) BCI. [Figure: Schematic of a hybrid fNIRS-EEG BCI.] Several modalities have been used for brain signal acquisition, which include EEG (Wolpaw et al., 2002; Turnip et al., 2011; Turnip and Hong, 2012; Wang et al., 2012; Hwang et al., 2013; Kleih and Kubler, 2013; Ko and Sim, 2013; Hammer et al., 2014; Kim et al., 2014; Soekadar et al., 2014), magnetoencephalography (MEG) (Mellinger et al., 2007; Buch et al., 2008; Sardouie and Shamsollahi, 2012), functional magnetic resonance imaging (fMRI) (Weiskopf et al., 2004; LaConte, 2011; van der Heiden et al., 2014), and fNIRS (Ferrari et al., 1985, 2004; Kato et al., 1993; Hu et al., 2013; Bhutta et al., 2014; Rea et al., 2014; Santosa et al., 2014). Among them, fNIRS is relatively new, which uses near-infrared-range light (usually of 650~1000 nm wavelength) to measure the concentration changes of oxygenated hemoglobin (HbO) and deoxygenated hemoglobin (HbR) (Villringer et al., 1993; Hoshi et al., 1994; Hoshi and Tamura, 1997; Villringer and Chance, 1997; Boas et al., 2004a,b; Hong and Nguyen, 2014). Its main advantages are relatively low cost, portability, safety, low noise (compared to fMRI), and easiness to use. Unlike EEG and MEG, its data are not much susceptible to electrical noise, since it is an optical imaging modality. fNIRS measures the blood flow changes in the local capillary network caused by neuron firings. Since the hemoglobin is an oxygen carrier, the changes of HbO and HbR concentration levels after a neuronal activation can be related to the relevant neuronal firings. fNIRS uses near-infrared (NI) light emitter-detector pairs operating with two or more wavelengths. The NI light emitted into the scalp diffuses through the brain tissues resulting in multiple scattering of photons. Some of these photons exit the head after passing through the cortical region of the brain, wherein the chromophores (i.e., HbO and HbR) are changing in time. These exited photons are then detected by using strategically positioned detectors. Since HbO and HbR have different absorption coefficients for different wavelengths of NI light, the relationship between the exiting-photon intensity and the incident-photon intensity can be used to calculate the changes of the concentrations of HbO and HbR [ΔcHbO(t) and ΔcHbR(t)] along the path of the photons by applying the modified Beer-Lamberts law (Delpy et al., 1988). The principle of fNIRS measurement, first reported by Jobsis (1977), has been applied to the study of cerebral hemodynamics for more than two decades, even though its BCI use is only a few years old. The first study who demonstrated the feasibility of fNIRS for BCI was Coyle et al. (2004). They asked the subjects to perform motor imagery of continuous squeezing and releasing of a soft ball. Based on the activity threshold of ΔcHbO(t), they determined whether the brain was activated or at rest. In 2007, three studies demonstrated the feasibility of controlling the output of fNIRS BCI: Coyle et al. (2007) used a custom-built fNIRS system (named Mindswitch) to test on-off control. Their protocol consisted of two options alternately presented to the subjects: When a desired option was highlighted, the subject performed motor imagery of squeezing and releasing a soft ball to enhance the HbO signals in the motor cortex and, in this way, expressed their choice mentally. The signals during motor imagery were classified against those during the rest period with an average accuracy of more than 80%. Sitaram et al. (2007) showed that fNIRS signal patterns during execution movement and imagery were distinguishable with the accuracy of 80% (or above) using support vector machines (SVM) and hidden Markov model (HMM). On the other hand, the first investigation on ALS patients was done by Naito et al. (2007): Forty ALS patients (17 of them were totally locked-in) were asked to encode their response to several questions as “yes” or “no.” They were requested to respond “yes” by performing mental calculation, music imagery and other such tasks, and to respond “no” by remaining relaxed. The instantaneous amplitude and phase of the light-intensity signals were then used as the features for a quadratic discriminant analysis classifier, which successfully decoded the responses

Read the complete text