Hendrik Santosa

Decoding four different sound-categories in the auditory cortex using functional near-infrared spectroscopy.

The ability of the auditory cortex in the brain to distinguish different sounds is important in daily life. This study investigated whether activations in the auditory cortex caused by different sounds can be distinguished using functional near-infrared spectroscopy (fNIRS). The hemodynamic responses (HRs) in both hemispheres using fNIRS were measured in 18 subjects while exposing them to four sound categories (English-speech, non-English-speech, annoying sounds, and nature sounds). As features for classifying the different signals, the mean, slope, and skewness of the oxy-hemoglobin (HbO) signal were used. With regard to the language-related stimuli, the HRs evoked by understandable speech (English) were observed in a broader brain region than were those evoked by non-English speech. Also, the magnitudes of the HbO signals evoked by English-speech were higher than those of non-English speech. The ratio of the peak values of non-English and English speech was 72.5%. Also, the brain region evoked by annoying sounds was wider than that by nature sounds. However, the signal strength for nature sounds was stronger than that for annoying sounds. Finally, for brain-computer interface (BCI) purposes, the linear discriminant analysis (LDA) and support vector machine (SVM) classifiers were applied to the four sound categories. The overall classification performance for the left hemisphere was higher than that for the right hemisphere. Therefore, for decoding of auditory commands, the left hemisphere is recommended. Also, in two-class classification, the annoying vs. nature sounds comparison provides a higher classification accuracy than the English vs. non-English speech comparison. Finally, LDA performs better than SVM.

The NIRS Brain AnalyzIR Toolbox

Functional near-infrared spectroscopy (fNIRS) is a noninvasive neuroimaging technique that uses low-levels of light (650–900 nm) to measure changes in cerebral blood volume and oxygenation. Over the last several decades, this technique has been utilized in a growing number of functional and resting-state brain studies. The lower operation cost, portability, and versatility of this method make it an alternative to methods such as functional magnetic resonance imaging for studies in pediatric and special populations and for studies without the confining limitations of a supine and motionless acquisition setup. However, the analysis of fNIRS data poses several challenges stemming from the unique physics of the technique, the unique statistical properties of data, and the growing diversity of non-traditional experimental designs being utilized in studies due to the flexibility of this technology. For these reasons, specific analysis methods for this technology must be developed. In this paper, we introduce the NIRS Brain AnalyzIR toolbox as an open-source Matlab-based analysis package for fNIRS data management, pre-processing, and first- and second-level (i.e., single subject and group-level) statistical analysis. Here, we describe the basic architectural format of this toolbox, which is based on the object-oriented programming paradigm. We also detail the algorithms for several of the major components of the toolbox including statistical analysis, probe registration, image reconstruction, and region-of-interest based statistics.

Decoding auditory information from the human brain using functional near infrared spectroscopy

Introduction T he present study reports the ability to decode auditory activation in the cortical region of the human brain evoked by multiple soundcategories for both offline and online classification. To that end, the haemodynamic response (HR) is measured in the leftand right-hemispheres of the auditory cortex using functional near infrared spectroscopy (fNIRS). In this article, six sound-categories (English-speech, non-English-speech, annoying sounds, nature sounds, classical music and gunshot sounds) are studied (25 subjects), and two-class (i.e., English vs non-English, annoying vs nature sounds), four-class (i.e., English, non-English, annoying and nature sounds) and six-class classification problems are discussed. The two wavelengths of the fNIRS used in this work are 760 nm and 830 nm; relevant light-absorbing chromophores are oxy-haemoglobin (HbO) and deoxy-haemoglobin (HbR), the concentrations of which are calculated using a modified Beer–Lambert law. fNIRS has several advantages over other modalities, including non-invasiveness, portability, safety and its low cost. Additionally, fNIRS offers a good trade-off between spatial and temporal resolutions compared with electroencephalography and functional magnetic resonance imaging (fMRI), respectively. Given these advantages, NIR techniques have been applied to various applications such as neurology, experimental psychology and brain research. A limitation of fNIRS, however, is the fact that the maximum light penetration depth is about 4 cm. It has an exceptional spatial resolution for monitoring the HR in the auditory cortex. Another benefit of this technique is its more silent optical neuroimaging compared with fMRI. Indeed, fMRI is relatively problematic for auditory cortex studies, as its associated mechanical noise can cause interference when measuring task-evoked brain activities. Overall, fNIRS is a very suitable technique for measuring the auditory cortex in the human brain and allows the subject either to sit in an upright position or lie down. The HRs (i.e., HbO and HbR) measured by fNIRS are combinations of various components: a task-evoked signal, a non-evoked signal (e.g., spontaneous fluctuations in about <0.01 Hz), physiological processes (e.g., respiratory and cardiac in about ~0.2– 0.3 Hz and ~1 Hz, respectively) and motion artefacts. In fNIRS data processing, such unwanted noise (excluding the task-evoked signal) needs to be reduced or removed in order to obtain higher-quality topographic images using adaptive filtering and/or the independent component analysis (ICA) algorithm. Further, the classification approach is a technique that uses feature selection to estimate data classes. For example, the mean, slope, skewness and kurtosis values of HbO and HbR signals for every trial over all channels were used as good features for classification. In this study, two classifiers in linear discriminant analysis (LDA) and support vector machines (SVMs), the most popular algorithms for classification, were applied to the multi-class problems. The aim of this study is to decode what humans are hearing by measuring the audiotask-evoked signal of the auditory activation bilaterally using fNIRS. Sounds from different sound-categories, selected from a website, were presented to 25 subjects. Pre-processing techniques (i.e., filtering), statistical analysis and the ICA algorithm were used for noise reduction in order to enhance classification accuracy. Further, the combinations of the mean, slope, skewness and kurtosis values of the HbO and HbR signals were calculated in the classification process as a features set. Finally, decoding of the multiple sound categories via multi-class LDA and SVM classifiers in offline and online schemes was investigated. This article summarises previously published work (with additional sound categories and on-line independent component analysis) to assess the use of functional near-infrared spectroscopy in decoding auditory activation for both off-line and on-line classification. Materials and methods A total of 25 subjects participated in this study: 18 subjects (28.11 ± 4.32 years) and 7 subjects (27.33 ± 2.93 years) for four and six different sound-categories, respectively. All of the subjects had normal hearing and no history of any neurological disorder. The subjects were asked to listen to various audio stimuli attentively and to guess, for each stimulus, which category was heard. The experiment was conducted in accordance with the ethical standards outlined in the latest Declaration of Helsinki. Audio stimuli consisted of six different sound-categories selected from the website www.youtube.com: two speech languages (English, non-English), two types of sounds (annoying sounds, nature sounds), classical music and gunshot sounds. Categories 1–2, 3–4, 1–4 and 1–6 were used in the two-class (speech), two-class (sound), four-class and six-class problems, respectively. Accordingly, Subjects 1–18 and 19–25 were, respectively, exposed to 24 doi: 10.1255/nirn.1608

Fast optical signals during median nerve stimulation by using continuous-wave NIRS

The purpose of the present study was to investigate the reliability of fast optical signal (FOS) using near-infrared spectroscopy (NIRS). In this study, we presented a preliminary study in the detection of FOS from 10 subjects during electrical median nerve stimulation. We used high sampling frequency using continuous-wave NIRS to detect the FOS related to the neural activity. Furthermore, the FOS has been found with latency between 100 ms to 200 ms after stimulus.

Current BCI technologies in brain engineering

Various neuroimaging modalities have appeared to acquire brain signals for developing a brain-computer interface (BCI). In this article, we review studies on different modalities including both invasive and non-invasive techniques for the implementation of BCIs, for brain signals detection, decoding, feature extraction, and classification. We discuss their advantages, disadvantages, and implementation issues to design of BCIs. Finally, we focus on fundamental principles, recent developments, and applications by using optical imaging as a promising modality for BCI applications.

An fNIRS-based BCI for mental arithmetic task using ICA

Functional near-infrared spectroscopy (fNIRS) is a non-invasive methods for acquiring neuronal activation coupled hemodynamic changes from the brain evoked by some particular stimuli. Recently, the study of fNIRS for implementation of brain-computer interface (BCI), for the acquisition and decoding of hemodynamic signals in the brain. In order to generate neuronal signals, arithmetic tasks were administered to eight subjects. This paper proposes an independent component analysis (ICA) to extract hemodynamic signals and unwanted signals. Significantly, the clearer hemodynamic response has been found after restoring the independent components from ICA which is indicated by increasing the t-value.

Non-lateralization with noise in the auditory cortex: An fNIRS study

Functional near-infrared spectroscopy (fNIRS) technique is used to measure concentration of hemodynamic changes (i.e., oxy- and deoxy-hemoglobin) in the human brain, while a subject is performing specific cognitive tasks. The advance brain functions related to music are thought to implicate uniquely human abilities, and it is known to have a strong tendency for hemispheric lateralization in the brain. The purpose of this paper is to investigate the effects of lateralization brain function in auditory cortex using fNIRS for music stimuli with background noise. It was found that the response to music with different background noise will affect different lateralization as indicated in the activation maps.

Asymmetry brain function in auditory cortex: A functional near-infrared spectroscopy study

Functional near-infrared spectroscopy (fNIRS) can measure the change of hemodynamic response, it enables to determine the concentration changes of oxy-hemoglobin and deoxy-hemoglobin. The aim in this paper is to investigate the forms of lateralization or asymmetry brain function in auditory cortex using fNIRS. This technique shows good promise for assessment of asymmetry functions in the auditory cortex.