An fNIRS-Based Feature Learning and Classification Framework to Distinguish Hemodynamic Patterns in Children Who Stutter

Diagnostic machine learning applications on clinical populations using functional near infrared spectroscopy: a review

Functional near-infrared spectroscopy (fNIRS) and its interaction with machine learning (ML) is a popular research topic for the diagnostic classification of clinical disorders due to the lack of robust and objective biomarkers. This review provides an overview of research on psychiatric diseases by using fNIRS and ML. Article search was carried out and 45 studies were evaluated by considering their sample sizes, used features, ML methodology, and reported accuracy. To our best knowledge, this is the first review that reports diagnostic ML applications using fNIRS. We found that there has been an increasing trend to perform ML applications on fNIRS-based biomarker research since 2010. The most studied populations are schizophrenia (n = 12), attention deficit and hyperactivity disorder (n = 7), and autism spectrum disorder (n = 6) are the most studied populations. There is a significant negative correlation between sample size (>21) and accuracy values. Support vector machine (SVM) and deep learning (DL) approaches were the most popular classifier approaches (SVM = 20) (DL = 10). Eight of these studies recruited a number of participants more than 100 for classification. Concentration changes in oxy-hemoglobin (ΔHbO) based features were used more than concentration changes in deoxy-hemoglobin (ΔHb) based ones and the most popular ΔHbO-based features were mean ΔHbO (n = 11) and ΔHbO-based functional connections (n = 11). Using ML on fNIRS data might be a promising approach to reveal specific biomarkers for diagnostic classification.

fNIRS Studies of Individuals with Speech and Language Impairment Underreport Sociodemographics: A Systematic Review

Functional near-infrared spectroscopy (fNIRS) is a promising tool for scientific discovery and clinical application. However, its utility depends upon replicable reporting. We evaluate reporting of sociodemographics in fNIRS studies of speech and language impairment and asked the following: (1) Do refereed fNIRS publications report participant sociodemographics? (2) For what reasons are participants excluded from analysis? This systematic review was preregistered with PROSPERO (CRD42022342959) and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses protocol. Searches in August 2022 included the terms: (a) fNIRS or functional near-infrared spectroscopy or NIRS or near-infrared spectroscopy, (b) speech or language, and (c) disorder or impairment or delay. Searches yielded 38 qualifying studies from 1997 to present. Eight studies (5%) reported at least partial information on race or ethnicity. Few studies reported SES (26%) or language background (47%). Most studies reported geographic location (100%) and gender/sex (89%). Underreporting of sociodemographics in fNIRS studies of speech and language impairment hinders the generalizability of findings. Replicable reporting is imperative for advancing the utility of fNIRS.

Functional near-infrared spectroscopy in pediatric clinical research: Different pathophysiologies and promising clinical applications

Over its 30 years of existence, functional near-infrared spectroscopy (fNIRS) has matured into a highly versatile tool to study brain function in infants and young children. Its advantages, amongst others, include its ease of application and portability, the option to combine it with electrophysiology, and its relatively good tolerance to movement. As shown by the impressive body of fNIRS literature in the field of cognitive developmental neuroscience, the method's strengths become even more relevant for (very) young individuals who suffer from neurological, behavioral, and/or cognitive impairment. Although a number of studies have been conducted with a clinical perspective, fNIRS cannot yet be considered as a truly clinical tool. The first step has been taken in this direction by studies exploring options in populations with well-defined clinical profiles. To foster further progress, here, we review several of these clinical approaches to identify the challenges and perspectives of fNIRS in the field of developmental disorders. We first outline the contributions of fNIRS in selected areas of pediatric clinical research: epilepsy, communicative and language disorders, and attention-deficit/hyperactivity disorder. We provide a scoping review as a framework to allow the highlighting of specific and general challenges of using fNIRS in pediatric research. We also discuss potential solutions and perspectives on the broader use of fNIRS in the clinical setting. This may be of use to future research, targeting clinical applications of fNIRS in children and adolescents.

Lower limb exoskeleton robot and its cooperative control: A review, trends, and challenges for future research

Effective control of an exoskeleton robot (ER) using a human-robot interface is crucial for assessing the robot's movements and the force they produce to generate efficient control signals. Interestingly, certain surveys were done to show off cutting-edge exoskeleton robots. The review papers that were previously published have not thoroughly examined the control strategy, which is a crucial component of automating exoskeleton systems. As a result, this review focuses on examining the most recent developments and problems associated with exoskeleton control systems, particularly during the last few years (2017-2022). In addition, the trends and challenges of cooperative control, particularly multi-information fusion, are discussed.

Looking for “fNIRS Signature” in Autism Spectrum: A Systematic Review Starting From Preschoolers

Accumulating evidence suggests that functional Near-Infrared Spectroscopy (fNIRS) can provide an essential bridge between our current understanding of neural circuit organization and cortical activity in the developing brain. Indeed, fNIRS allows studying brain functions through the measurement of neurovascular coupling that links neural activity to subsequent changes in cerebral blood flow and hemoglobin oxygenation levels. While the literature offers a multitude of fNIRS applications to typical development, only recently this tool has been extended to the study of neurodevelopmental disorders (NDDs). The exponential rise of scientific publications on this topic during the last years reflects the interest to identify a “fNIRS signature” as a biomarker of high translational value to support both early clinical diagnosis and treatment outcome. The purpose of this systematic review is to describe the updating clinical applications of fNIRS in NDDs, with a specific focus on preschool population. Starting from this rationale, a systematic search was conducted for relevant studies in different scientific databases (Pubmed, Scopus, and Web of Science) resulting in 13 published articles. In these studies, fNIRS was applied in individuals with Autism Spectrum Disorder (ASD) or infants at high risk of developing ASD. Both functional connectivity in resting-state conditions and task-evoked brain activation using multiple experimental paradigms were used in the selected investigations, suggesting that fNIRS might be considered a promising method for identifying early quantitative biomarkers in the autism field.

Consistency of children's hemodynamic responses during spontaneous speech

Significance: Hemodynamic responses (HRs) are typically averaged across experimental sessions based on the assumption that brain activation is consistent over multiple trials. This may not be a safe assumption, especially in pediatric populations, due to unaccounted effects of inattention, fatigue, or habituation. Aim: The purpose of this study was to quantify the consistency of the HR over speech and language brain regions during speech production in typically developing school-aged children. Approach: Brain activity over speech and language regions of interest (ROIs) was recorded with functional near-infrared spectroscopy during a picture description paradigm with 37 children (aged 7 to 12 years). We divided the 30 experimental trials, each 5 s long, into three segments of 10 trials each corresponding with early (trials 1 to 10), middle (trials 11 to 20), and late (trials 21 to 30) trials. We then compared oxygenated (HbO) and deoxygenated (HbR) hemoglobin concentrations averaged across each 10 trial segment to overall concentrations averaged across all 30 trials. We also compared differential hemoglobin (HbD) across ROIs. Results: HbO and HbR averaged across all experimental trials most strongly correlated with HbO and HbR from early trials. HbD values from channels over most speech and language regions did not appreciably change throughout the experimental session. The exception was HbD values from channels over the dorsal inferior frontal gyrus (dIFG). This region showed significantly higher activation over the left hemisphere during the first segment of the experiment. Conclusions: Our findings suggest that brain activity from speech and language ROIs was relatively consistent over the experimental session. The exception was increased activation of left dIFG during earlier experimental trials. We suggest that researchers critically evaluate the consistency of HRs from different brain regions to determine the reliability of HRs recorded during experimental sessions. This step is instrumental in ensuring that uncontrolled effects do not mask patterns of task-related activation.

Crossing time windows optimization based on mutual information for hybrid BCI

Hybrid EEG-fNIRS brain-computer interface (HBCI) is widely employed to enhance BCI performance. EEG and fNIRS signals are combined to increase the dimensionality of the information. Time windows are used to select EEG and fNIRS singles synchronously. However, it ignores that specific modal signals have their own characteristics, when the task is stimulated, the information between the modalities will mismatch at the moment, which has a significant impact on the classification performance. Here we propose a novel crossing time windows optimization for mental arithmetic (MA) based BCI. The EEG and fNIRS signals were segmented separately by sliding time windows. Then crossing time windows (CTW) were combined with each one segment from EEG and fNIRS selected independently. Furthermore, EEG and fNIRS features were extracted using Filter Bank Common Spatial Pattern (FBCSP) and statistical methods from each sample. Mutual information was calculated for FBCSP and statistical features to characterize the discrimination of crossing time windows, and the optimal window would be selected based on the largest mutual information. Finally, a sparse structured framework of Fisher Lasso feature selection (FLFS) was designed to select the joint features, and conventional Linear Discriminant Analysis (LDA) was employed to perform classification. We used proposed method for a MA dataset. The classification accuracy of the proposed method is 92.52 ± 5.38% and higher than other methods, which shows the rationality and superiority of the proposed method.

Exploring Relationships Among Risk Factors for Persistence in Early Childhood Stuttering

Purpose The purpose of this study is to investigate how epidemiological and clinical factors collectively predict whether a preschooler who is stuttering will persist or recover and to provide guidance on how clinicians can use these factors to evaluate a child's risk for stuttering persistence. Method We collected epidemiological and clinical measures from 52 preschoolers (M = 54.4 months, SD = 6.7 months; 38 boys and 14 girls) diagnosed as stuttering. We then followed these children longitudinally to document whether they eventually recovered or persisted in stuttering. Risk factors found to be significantly associated with stuttering persistence were used to build single and multiple variable predictive statistical models. Finally, we assessed each model's prediction capabilities by recording how accurate a model was in predicting a child's stuttering outcome-persisting or recovered. Results We found that a positive family history of stuttering, poorer performance on a standardized articulation/phonological assessment, higher frequency of stuttering-like disfluencies during spontaneous speech, and lower accuracy on a nonword repetition task were all significantly associated with an increased probability of persistence. The interaction between family history of stuttering and nonword repetition performance was also significant. The full multiple regression model incorporating all these risk factors resulted in the best fitting model with the highest predictive accuracy and lowest error rate. Conclusions For the first time, we show how multiple risk factors collectively predict the probability of stuttering persistence in 3- to 5-year-old preschool children who stutter. Using the full combination of risk factors to assess preschoolers who stutter yielded more accurate predictions of persistence compared to sparser models. A better understanding of the factors that underlie stuttering persistence will yield insight into the underpinnings of chronic stuttering and will help identify etiological targets for novel treatment approaches.

Predictive factors for persistence and recovery of stuttering in children: A systematic review

PURPOSE

The purpose of this study was to systematically review the available literature on various factors that can predict the persistence and recovery of stuttering in children.

METHOD

An electronic search yielded a total of 35 studies, which considered 44 variables that can be potential factors for predicting persistence and recovery.

RESULT

Among 44 factors studied, only four factors- phonological abilities, articulatory rate, change in the pattern of disfluencies, and trend in stuttering severity over one-year post-onset were identified to be replicated predictors of recovery of the stuttering. Several factors, such as differences in the second formant transition between fluent and disfluent speech, articulatory rate measured in phones/sec, etc., were observed to predict the future course of stuttering. However, these factors lack replicated evidence as predictors.

CONCLUSION

There is clear support only for limited factors as reliable predictors. Also, it is observed to be too early to conclude on several replicated factors due to differences in the age group of participants, participant sample size, and the differences in tools used in research that lead to mixed findings as a predictive factor. Hence there is a need for systematic and replicated testing of the factors identified before initiating their use for clinical purposes.

Tracking Brain Development From Neonates to the Elderly by Hemoglobin Phase Measurement Using Functional Near-Infrared Spectroscopy

The biological and neurological processes during the lifespan are dynamic with significant alterations associated with different stages of life. The phase and coupling of oxy-hemoglobin (Δ[HbO]) and deoxy-hemoglobin concentration changes (Δ[Hb]) measured by functional near-infrared spectroscopy (fNIRS) are shown to characterize the neurovascular and metabolic development of infants. However, the changes in phase and coupling across the human lifespan remain mostly unknown. Here, fNIRS measurements of Δ[HbO] and Δ[Hb] conducted at two sites on different age populations (from newborns to elderly) were combined. Firstly, we assessed the influence of random noise on the calculation of the phase difference and phase-locking index (PLI) in fNIRS measurement. The results showed that the phase difference is close to π as the noise intensity approaches -8 dB, and the coupling strength (i.e., PLI) presents a u-shape curve as the noise increase. Secondly, phase difference and PLI in the frequency range 0.01-0.10 Hz were calculated after denoising. It showed that the phase difference increases from newborns to 3-4-month-olds babies. This phase difference persists throughout adulthood until finally being disrupted in the old age. The children's PLI is the highest, followed by that of adults. These two groups' PLI are significantly higher than those of infants and the elderly (p < 0.001). Lastly, a hemodynamic model was used to explain the observations and found close associations with cerebral autoregulation and speed of blood flow. These results demonstrate that the phase-related parameters measured by fNIRS can be used to study the brain and assess brain health throughout the lifespan.

Motor Training Using Mental Workload (MWL) With an Assistive Soft Exoskeleton System: A Functional Near-Infrared Spectroscopy (fNIRS) Study for Brain-Machine Interface (BMI)

Mental workload is a neuroergonomic human factor, which is widely used in planning a system's safety and areas like brain-machine interface (BMI), neurofeedback, and assistive technologies. Robotic prosthetics methodologies are employed for assisting hemiplegic patients in performing routine activities. Assistive technologies' design and operation are required to have an easy interface with the brain with fewer protocols, in an attempt to optimize mobility and autonomy. The possible answer to these design questions may lie in neuroergonomics coupled with BMI systems. In this study, two human factors are addressed: designing a lightweight wearable robotic exoskeleton hand that is used to assist the potential stroke patients with an integrated portable brain interface using mental workload (MWL) signals acquired with portable functional near-infrared spectroscopy (fNIRS) system. The system may generate command signals for operating a wearable robotic exoskeleton hand using two-state MWL signals. The fNIRS system is used to record optical signals in the form of change in concentration of oxy and deoxygenated hemoglobin (HbO and HbR) from the pre-frontal cortex (PFC) region of the brain. Fifteen participants participated in this study and were given hand-grasping tasks. Two-state MWL signals acquired from the PFC region of the participant's brain are segregated using machine learning classifier-support vector machines (SVM) to utilize in operating a robotic exoskeleton hand. The maximum classification accuracy is 91.31%, using a combination of mean-slope features with an average information transfer rate (ITR) of 1.43. These results show the feasibility of a two-state MWL (fNIRS-based) robotic exoskeleton hand (BMI system) for hemiplegic patients assisting in the physical grasping tasks.

Effective brain connectivity for fNIRS data analysis based on multi-delays symbolic phase transfer entropy

OBJECTIVE

Recently, effective connectivity (EC) calculation methods for functional near-infrared spectroscopy (fNIRS) data mainly face two problems: the first problem is that noise can seriously affect the EC calculation and even lead to false connectivity; the second problem is that it ignores the various real neurotransmission delays between the brain region, and instead uses a fixed delay coefficient for calculation.

APPROACH

To overcome these two issues, a delay symbolic phase transfer entropy (dSPTE) is proposed by developing traditional transfer entropy (TE) to estimate EC for fNIRS. Firstly, the phase time sequence was obtained from the original sequence by the Hilbert transform and state-space reconstruction was realized using a uniform embedding scheme. Then, a symbolization technique was applied based on a neural-gas algorithm to improve its noise robustness. Finally, the EC was calculated on multiple time delay scales to match different inter-region neurotransmission delays.

MAIN RESULTS

A linear AR model, a nonlinear model and a multivariate hybrid model were introduced to simulate the performance of dSPTE, and the results showed that the accuracy of dSPTE was the highest, up to 74.27%, and specificity was 100% which means no false connectivity. The results confirmed that the dSPTE method realized better noise robustness, higher accuracy, and correct identification even if there was a long delay between series. Finally, we applied dSPTE to fNIRS dataset to analyse the EC during the finger-tapping task, the results showed that EC strength of task state significantly increased compared with the resting state.

SIGNIFICANCE

The proposed dSPTE method is a promising way to measure the EC for fNIRS. It incorporates the phase information TE with a symbolic process for fNIRS analysis for the first time. It has been confirmed to be noise robust and suitable for the complex network with different coupling delays.

Functional Near-Infrared Spectroscopy in the Study of Speech and Language Impairment Across the Life Span: A Systematic Review

Purpose Functional brain imaging is playing an increasingly important role in the diagnosis and treatment of communication disorders, yet many populations and settings are incompatible with functional magnetic resonance imaging and other commonly used techniques. We conducted a systematic review of neuroimaging studies using functional near-infrared spectroscopy (fNIRS) with individuals with speech or language impairment across the life span. We aimed to answer the following question: To what extent has fNIRS been used to investigate the neural correlates of speech-language impairment? Method This systematic review was preregistered with PROSPERO, the international prospective register of systematic reviews (CRD42019136464). We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol for preferred reporting items for systematic reviews. The database searches were conducted between February and March of 2019 with the following search terms: (a) fNIRS or functional near-infrared spectroscopy or NIRS or near-infrared spectroscopy, (b) speech or language, and (c) disorder or impairment or delay. Results We found 34 fNIRS studies that involved individuals with speech or language impairment across nine categories: (a) autism spectrum disorders; (b) developmental speech and language disorders; (c) cochlear implantation and deafness; (d) dementia, dementia of the Alzheimer's type, and mild cognitive impairment; (e) locked-in syndrome; (f) neurologic speech disorders/dysarthria; (g) stroke/aphasia; (h) stuttering; and (i) traumatic brain injury. Conclusions Though it is not without inherent challenges, fNIRS may have advantages over other neuroimaging techniques in the areas of speech and language impairment. fNIRS has clinical applications that may lead to improved early and differential diagnosis, increase our understanding of response to treatment, improve neuroprosthetic functioning, and advance neurofeedback.

Using a Data-Driven Approach to Estimate Second-Language Proficiency From Brain Activation: A Functional Near-Infrared Spectroscopy Study

While non-invasive brain imaging has made substantial contributions to advance human brain science, estimation of individual state is becoming important to realize its applications in society. Brain activations were used to classify second-language proficiencies. Participants in functional near-infrared spectroscopy (fNIRS) experiment were 20/20 native Japanese speakers with high/low English abilities and 19/19 native English speakers with high/low Japanese abilities. Their cortical activities were measured by functional near-infrared spectroscopy while they were conducting Japanese/English listening comprehension tests. The data-driven method achieved classification accuracy of 77.5% in the case of Japanese speakers and 81.9% in the case of English speakers. The informative features predominantly originated from regions associated with language function. These results bring an insight of fNIRS neuroscience and its applications in society.

Enhanced Accuracy for Multiclass Mental Workload Detection Using Long Short-Term Memory for Brain-Computer Interface

Cognitive workload is one of the widely invoked human factors in the areas of human-machine interaction (HMI) and neuroergonomics. The precise assessment of cognitive and mental workload (MWL) is vital and requires accurate neuroimaging to monitor and evaluate the cognitive states of the brain. In this study, we have decoded four classes of MWL using long short-term memory (LSTM) with 89.31% average accuracy for brain-computer interface (BCI). The brain activity signals are acquired using functional near-infrared spectroscopy (fNIRS) from the prefrontal cortex (PFC) region of the brain. We performed a supervised MWL experimentation with four varying MWL levels on 15 participants (both male and female) and 10 trials of each MWL per participant. Real-time four-level MWL states are assessed using fNIRS system, and initial classification is performed using three strong machine learning (ML) techniques, support vector machine (SVM), k-nearest neighbor (k-NN), and artificial neural network (ANN) with obtained average accuracies of 54.33, 54.31, and 69.36%, respectively. In this study, novel deep learning (DL) frameworks are proposed, which utilizes convolutional neural network (CNN) and LSTM with 87.45 and 89.31% average accuracies, respectively, to solve high-dimensional four-level cognitive states classification problem. Statistical analysis, t-test, and one-way F-test (ANOVA) are also performed on accuracies obtained through ML and DL algorithms. Results show that the proposed DL (LSTM and CNN) algorithms significantly improve classification performance as compared with ML (SVM, ANN, and k-NN) algorithms.

What Are Predictors for Persistence in Childhood Stuttering?

Over the past 10 years, we (the Purdue Stuttering Project) have implemented longitudinal studies to examine factors related to persistence and recovery in early childhood stuttering. Stuttering develops essentially as an impairment in speech sensorimotor processes that is strongly influenced by dynamic interactions among motor, language, and emotional domains. Our work has assessed physiological, behavioral, and clinical features of stuttering within the motor, linguistic, and emotional domains. We describe the results of studies in which measures collected when the child was 4 to 5 years old are related to eventual stuttering status. We provide supplemental evidence of the role of known predictive factors (e.g., sex and family history of persistent stuttering). In addition, we present new evidence that early delays in basic speech motor processes (especially in boys), poor performance on a nonword repetition test, stuttering severity at the age of 4 to 5 years, and delayed or atypical functioning in central nervous system language processing networks are predictive of persistent stuttering.