Next Article in Journal
Leveraging a Repetitive Filtering Framework for Robust Pulse Stream Extraction in Distributed Optical Fiber Sensing Signals
Previous Article in Journal
Demonstration of Real-Time 4 × 89 km Core-Division-Multiplexed Transmission Using Weakly Coupled Seven-Core Fiber and C+L-Band 1.2 Tb/s OTN Transponders with Configurable Baud Rate
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Photonics
Volume 13
Issue 1
10.3390/photonics13010054
photonics-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Functional Near-Infrared Spectroscopy in Linguistic Research: Recent Advances and Future Perspectives

by
Pengke Cui
1,
Yezhi Cui
1,
Xin Zhang
2 and
Xiu Zhang
2,*
1
School of Foreign Languages, Tianjin University of Technology and Education, Tianjin 300222, China
2
School of Electronic and Communication Engineering, Tianjin Normal University, Tianjin 300072, China
*
Author to whom correspondence should be addressed.
Photonics 2026, 13(1), 54; https://doi.org/10.3390/photonics13010054
Submission received: 18 October 2025 / Revised: 24 December 2025 / Accepted: 30 December 2025 / Published: 7 January 2026
(This article belongs to the Section Biophotonics and Biomedical Optics)
Browse Figures
Versions Notes

Abstract

Functional Near-Infrared Spectroscopy (fNIRS), a non-invasive neuroimaging technique, has demonstrated unique advantages in linguistic research in recent years. By monitoring changes in the concentrations of oxygenated and deoxygenated hemoglobin during cortical activation, fNIRS provides new insights into the mechanisms underlying language processing. Its ecological validity and high compatibility enable seamless integration into real-world environments, minimizing interference and ensuring the authenticity of the collected data. In the realm of linguistics, fNIRS has been applied to studies on language perception, function, acquisition, cross-linguistic processing, and the assessment of language disorders, revealing the intricate mechanisms of language processing and showcasing its potential for clinical applications. This article reviews the latest advancements in the utilization of fNIRS in linguistic research, aiming to provide valuable references for researchers and to foster deeper exploration and innovative development in this field. Meanwhile, this article systematically examines the limitations of fNIRS in current research, provides a critical assessment of its methodological and applicative value, and, on this basis, outlines future directions and potential breakthroughs for this technology in the field of language research.

1. Introduction

Functional Near-Infrared Spectroscopy (fNIRS) is a technique that utilizes near-infrared light to measure changes in the concentrations of oxygenated hemoglobin (HbO) and deoxygenated hemoglobin (HbR) within the cerebral cortex [1,2]. This technology is portable, wearable, non-invasive, and can even be powered by batteries and connected wirelessly, rendering it particularly suitable for brain–computer interface interactions in everyday environments [3,4]. Depending on anatomical differences in head structure, particularly the scalp-to-cortex distance, which is generally shorter in infants and longer in adults, this measure serves as the primary factor in determining the source-detector (SD) separation, while head circumference is considered only as a supplementary reference [5,6]. Compared to other neuroimaging techniques, fNIRS exhibits higher ecological validity, enabling seamless integration into natural environments, minimizing interference, and ensuring the authenticity of the collected data. These characteristics confer unique advantages to fNIRS in studies of language perception, function, acquisition, and cross-linguistic processing.
The pioneering application of near-infrared spectroscopy in humans dates back to 1977. Initially, this technology was primarily employed in neonatal and adult populations, focusing on experimental exploration and clinical monitoring of cerebral oxygenation [7,8]. Subsequent empirical studies by multiple research teams demonstrated that near-infrared spectroscopy could not only penetrate the intact adult skull to assess brain activation but also possessed unique suitability for neonatal brain function research due to its non-invasive and low-interference characteristics [9,10,11]. To date, fNIRS has been widely adopted across various fields, including cognitive neuroscience, educational and learning sciences, and language research. Several high-quality reviews have provided comprehensive overviews of the advancements in these domains [12,13,14,15]. Building upon existing research, this article conducts a systematic review of the application of fNIRS in linguistics, with a focus on key areas such as speech perception and recognition, language lateralization, language acquisition, cross-linguistic studies, clinical applications and applications of multimodal neuroimaging techniques. This structure follows a logical progression from fundamental processes to developmental patterns, clinical translation, and technical integration. Furthermore, it thoroughly summarizes the current limitations of the technology and discusses future prospects for its use in linguistic research. Speech production and perception primarily rely on Broca’s area, Wernicke’s area, Exner’s area, and the Visual Word Form area; their relative positions are shown in the simplified schematic diagram in Figure 1.

2. fNRIS Methodology

The core physical foundation of fNIRS functional imaging relies on the propagation characteristics of near-infrared light in biological tissues and the differences in optical absorption between hemoglobin species. First, biological tissues (such as the scalp, skull, and cortex) exhibit a “therapeutic optical window” within the 700–900 nm wavelength range; near-infrared light in this band is weakly absorbed by water and fat, allowing for effective penetration into cortical tissues at depths of 1.5–2 cm, thereby enabling the detection of brain functional signals [16,17]. Second, HbO and HbR possess distinct absorption spectra, with an “isosbestic point” at approximately 805 nm, a wavelength at which their molar extinction coefficients are equal [18]. By measuring the differences in light intensity attenuation at this wavelength and other characteristic wavelengths (e.g., 730 nm and 850 nm), changes in the concentrations of HbO and HbR can be calculated based on the Lambert-Beer law. This, in turn, indirectly reflects local hemodynamic alterations induced by neural activity in the cortex [19,20].
Table 1 delineates the operational principles and core characteristics of three primary fNIRS methodologies, each tailored to distinct application needs in biomedical monitoring. Time-Domain fNIRS (TD-fNIRS) systems employ ultra-short picosecond-level pulsed light to directly capture the temporal distribution of photon time-of-flight within biological tissues. This unique capability enables the precise separation of optical absorption and scattering effects, and its time-gating technology delivers unrivaled depth resolution for layered tissue oxygenation monitoring, making it the gold standard for quantitative optical parameter measurement. However, this performance comes at the cost of extremely high system complexity, requiring specialized components such as time-correlated single photon counters, along with prohibitive hardware expenses and limited portability [21,22]. Frequency-Domain fNIRS (FD-fNIRS) systems, by comparison, utilize radio-frequency amplitude-modulated continuous light to detect two critical optical signatures: phase shift and amplitude attenuation of emergent photons. By solving the complex form of the photon diffusion equation, FD-fNIRS achieves reliable quantification of absolute hemoglobin concentrations and reduced scattering coefficients, striking a robust compromise between performance and practicality. It offers superior depth resolution and anti-interference ability over continuous-wave systems, with moderate complexity and cost that render it suitable for clinical quantitative monitoring scenarios [21]. In contrast, Continuous-Wave (CW-fNIRS) systems represent the most prevalent and cost-effective technology in the fNIRS family [21,23]. Operating with constant-intensity near-infrared light, they measure only the attenuation of emergent light intensity and rely on the Lambert-Beer Law to derive relative changes in hemoglobin concentrations. While they excel in real-time dynamic monitoring with high sampling rates, compact size, and low cost, they cannot inherently disentangle absorption and scattering coefficients, leading to limited quantitative accuracy and poor depth resolution. A fundamental commonality among all three modalities is their shared reliance on the hemodynamic response to indirectly infer underlying neural function in both basic research and clinical settings [21].
Currently, fNIRS research in the field of language almost entirely relies on two commercial continuous-wave (CW) systems: the Hitachi ETG-100/4000 (with wavelengths of 695/830 nm, 20 channels, a 3 cm inter-optode distance, and a sampling rate of 10 Hz) and the NIRx NIRScout (with wavelengths of 760/850 nm, a 16 × 16 optode array, 52 channels, and a sampling rate of 7.8 Hz). Given the complexity of Frequency-Domain (FD) and Time-Domain (TD) systems, along with their cumbersome pre-calibration procedures and stringent signal-to-noise ratio requirements, it is challenging to integrate them into everyday language interaction scenarios. To date, there have been no empirical reports on these two modalities in this context.

3. Literature Review

To systematically summarize the latest advancements in the application of fNIRS in linguistic research, this review applied the following four literature selection criteria: (1) Topical Relevance: Studies were included if they focused on applying fNIRS to core areas of language processing (e.g., speech perception, language acquisition, cross-linguistic comparison). Studies that used fNIRS only as an auxiliary tool without an in-depth exploration of linguistic mechanisms were excluded. (2) Publication Date Range: The search was limited to studies published between 2018 and 2024. (3) Language Restriction: Only literature written in Chinese or English was included, as the authors are proficient in these languages, and the core findings in current fNIRS linguistic research are primarily published in journals using them. (4) Research Quality Requirement: Non-full-text literature, such as abstracts and conference summaries, was excluded, with priority given to peer-reviewed journal articles.

3.1. Speech Perception and Recognition

This section reviews fNIRS studies on adult and infant speech perception, audiovisual integration, and speech processing under noise, revealing the cooperative roles of classic language regions and the motor cortex in spoken-language processing.
Early research using fNIRS focused on identifying the classical speech-related brain regions involved in language perception. These foundational studies demonstrated that near-infrared imaging could effectively detect cortical activation within these well-known areas, enhancing our fundamental understanding of the neural basis of speech perception. For example, in 2021, a team led by Ming Dong at Tianjin University conducted a study involving 13 subjects who repeated words overtly and covertly, recording fNIRS signals from the bilateral motor and prefrontal cortices. The study investigated specific brain regions and their corresponding cortico-cortical functional connectivity characteristics during imagined speech, providing fNIRS evidence for the involvement of the dorsal sensorimotor cortex in imagined speech. This research revealed that not only were traditional language areas (such as Broca’s area and Wernicke’s area) significantly active during imagined speech, but also that the motor cortex and prefrontal cortex exhibited active neural signals, indicating complex collaboration among these regions in language processing [24].
As the field progressed, researchers began to explore more complex language processing scenarios, such as audiovisual integration. At the Vanderbilt Brain Institute in the United States, a team led by Iliza M. Butera measured the extent to which 24 normal-hearing individuals integrated the audio of mono-osyllabic words with the corresponding visual signals from a female speaker at −6 and −9 dB signal-to-noise ratios. Using fNIRS, this study is the first to establish both a behavioral paradigm and cortical markers for audiovisual speech integration under noisy conditions in normal-hearing adults, thereby laying the groundwork for future investigations of how cochlear-implant users rely on this mechanism [25].
Building on the understanding of language processing under normal and audiovisual conditions, researchers then turned their attention to the impact of noise on speech recognition. At the University of Tennessee Health Science Center, Jessica Defenderfer’s team used fNIRS to study cortical responses in normal-hearing adults. Participants repeated vocoded or noise-masked sentences under high/low-intelligibility conditions (quiet sentences as baseline). With a new fNIRS image-reconstruction method, they found low-intelligibility speech activated the middle temporal gyrus (MTG) and middle frontal gyrus (MFG). MTG activity was specific to vocoded speech, MFG activity was specific to noise-masked speech, and MFG activation was higher during correct recognition. This shows untrained listeners use different frontal attentional mechanisms for vocoded and naturally degraded speech, relying on temporal-lobe analysis for vocoded stimuli [26].
The exploration of language perception and recognition using fNIRS also extended to the earliest stages of human life. Human neonates can discriminate between phonemes, yet the neural mechanisms underlying this ability remain unclear. In 2022, a research team led by Dandan Zhang at the School of Psychology, Shenzhen University, employed fNIRS to track neuroplastic changes in newborns exposed to front and back vowels. Neural responses were recorded at T1, within five hours after birth during the first random exposure, and again at T2, two hours later. Infants in the experimental group received identical stimulation during both T1 and T2, whereas controls received no training. Compared with controls, trained neonates exhibited shorter hemodynamic response latencies to front versus back vowels at T1, with the maximal difference localized in the inferior frontal gyrus. At T2, the differential neural activity increased further, becoming most pronounced in the superior temporal and left inferior parietal regions. These findings demonstrate that newborns undergo ultra-rapid tuning to natural phonemes within the first hours of life [27].
In 2024, the team led by Hewen Zhang utilized fNIRS signals from various brain regions to investigate the accuracy of task state detection across these regions. The results demonstrated that utilizing multiple channels within Broca’s area could achieve a task state detection accuracy comparable to that obtained by utilizing all channels across the entire brain. This finding implies that specific brain regions play a pivotal role in task state detection, and that optimizing the acquisition and analysis of fNIRS signals holds the potential to enhance the performance of brain–computer interfaces (BCIs) and neurofeedback systems [28].

3.2. Language Lateralization

Language lateralization is a prominent aspect of brain lateralization, referring primarily to the division and differences in language functions between the left and right hemispheres. Since Broca’s discovery in the mid-19th century that the brain region primarily responsible for language production is situated in the left inferior frontal gyrus, scientists have conducted extensive research on it. Using data from infants to adults, this part summarizes how fNIRS has been employed to investigate hemispheric specialization for language, particularly the emergence and maintenance of left-hemisphere dominance.
The exploration of language lateralization across different age groups reveals a fascinating developmental trajectory. In 2022, Fen Zhang’s team at Ghent University in Belgium employed fNIRS to investigate early language processing abilities in infants. The study involved 78 infants aged between 5 and 10 months, examining their hemodynamic responses in the temporal cortex in response to forward and backward speech in their native language, Dutch. The results showed that 5-month-old infants exhibited bilateral activation in response to both forward and backward speech, with no significant hemispheric lateralization for language. However, by 10 months, the left cortical response to forward speech became more pronounced. This indicates that significant developmental changes occur in the neural mechanisms underlying language processing during the first year of life, likely due to the accumulation of linguistic experience and the maturation of brain structures [29].
Building on this understanding of early-stage language lateralization development, researchers then shifted their focus to adults to further explore the established patterns of hemispheric specialization. In 2024, Hai-Jing Niu’s team at Beijing Normal University employed multichannel fNIRS to examine the hemispheric lateralization of visuospatial attention and language production in 52 healthy right-handed adults. During a picture-naming task, activation was strongly left-lateralized in the inferior frontal gyrus, whereas a landmark line-bisection task elicited right-lateralized activation in the inferior and superior parietal cortices. Quantitative laterality indices confirmed this double dissociation, and, critically, no correlation was found between the lateralization strengths of the two tasks, supporting the independence of these functional asymmetries [30].

3.3. Language Acquisition

Focusing on infants through adolescents, this section highlights fNIRS research on multimodal speech processing, emerging language networks, and neural mechanisms for tonal and phonological learning during language development.
Compared to fMRI, fNIRS is non-magnetic and portable, avoiding stress responses in infants caused by claustrophobia or physical restraint, and is adaptable to natural language learning contexts (e.g., parent–child interactive tasks), thereby improving ecological validity [31,32]. In contrast to EEG, fNIRS is less sensitive to scalp electrical interference, allowing more precise capture of cortical hemodynamic changes during language acquisition, which is particularly suitable for long-term tracking of infant brain development [31]. These features enable fNIRS to obtain data on the neural mechanisms of infant language learning in natural settings that are difficult to acquire with other technologies [32].
In 2018, Nicole Altvater-Mackensen from the Max Planck Institute for Human Cognitive and Brain Sciences in Germany investigated the non-modal recruitment of the inferior frontal cortex in human infants during speech processing. Using fNIRS, the study found that the speech-sensitive regions of the inferior frontal cortex in 6-month-old infants were activated regardless of whether they processed unimodal or multimodal speech stimuli. This suggests the inferior frontal cortex’s pivotal role in early language learning and speech perception, especially in integrating speech information across modalities [33]. Building on this, in 2019, Xuyun Wen’s team at the University of North Carolina at Chapel Hill took a step further. They utilized a modular information-guided approach combined with individual functional connectivity networks to construct robust, temporally consistent, and module-structure-enhanced group-level networks in infants. Their study, examining changes at three-month intervals, showed that functional brain network subdivisions became more refined over time, indicating enhanced functional segregation and integration in the first year of life. This reconfiguration improved information exchange efficiency, providing insights into the spatiotemporal changes in functional brain networks and laying the groundwork for understanding infant brain development [34].
In 2021, Rachael J. Lawrence’s team at the Nottingham Biomedical Research Centre, National Institute for Health Research (NIHR), United Kingdom, used fNIRS to measure bilateral temporal and frontal cortical activity in 6–13-year-old children with normal hearing while they listened to sentences presented at four different intelligibility levels, encoded in both clear and noisy conditions. The results indicated that, with increased speech clarity, the left temporal and frontal cortices exhibited stronger activation. This suggested the critical role of these regions in speech processing and provided a basis for developing new speech assessment and therapeutic methods [35].
The research on children’s speech processing also has implications for understanding the neural mechanisms of language production. In 2023, Jia-Nan Yu’s Master’s thesis conducted two fNIRS experiments using a picture–word interference paradigm to investigate the functional roles of tone and segment (i.e., initial consonant and vowel) in Chinese language production and their underlying neural mechanisms. The results showed significant overlap in neural activation between tone and segmental information in Chinese, with all (tone, initial consonants, and vowels) exhibiting marked left-brain dominance during processing, deepening our understanding of the neural basis of Chinese language processing [36]. Moreover, Martínez-Álvarez et al. combined fNIRS with behavioral paradigms to demonstrate that prosodic cues critically support early language learning in infants. Under prosodic highlighting, infants showed better acquisition and discrimination of nonadjacent rules, indicating that prosody provides additional cues to help infants extract and process complex grammatical information by enhancing their attention to target grammatical units [37].
In 2024, Hai-Jing Niu’s team at Beijing Normal University explored the neural processes involved in speech processing in infants, focusing specifically on the left frontal brain regions and hemispheric lateralization associated with the acquisition of Mandarin tones by Mandarin-speaking infants. The results revealed that the ability to discriminate tones improved with age, with older infants showing greater involvement of the frontal regions. This indicated the development of abstract tonal representations and an increase in bilateral activation, patterns similar to those observed in adult native Mandarin speakers. These findings contribute to a more comprehensive understanding of the relationship between native language acquisition and infant brain development during the critical period of early language learning [38]. Furthermore, Xin Zhou and colleagues’ research on how infant-directed speech (IDS) facilitates toddlers’ word learning can be linked to the above studies on infants’ neural responses. Compared with adult-directed speech, IDS elicited significantly stronger activation in the left dorsolateral prefrontal cortex, and neural responses within this region were positively correlated with individual vocabulary-learning performance. This indicates that IDS enhances attentional mechanisms, promoting lexical acquisition in young children, which is in line with the idea that various language-related cues and inputs play crucial roles in infants’ and toddlers’ language development [39].

3.4. Cross-Linguistic Studies

This segment synthesizes fNIRS findings on second-language processing, bilingual syntactic/semantic operations, cross-linguistic transfer, and inter-brain synchrony during dialogue, delineating both language-specific and shared neural signatures.
Leveraging its high ecological validity and capability to localize brain activation, fNIRS offers a unique perspective for cross-linguistic research, enabling real-time capture of brain activity during natural language tasks (e.g., conversation, language switching). This helps address core questions such as whether different languages share neural mechanisms and how bilingual experience reshapes brain networks. By overcoming the limitations of traditional imaging techniques in natural settings, fNIRS advances the study of language processing mechanisms from static monolingual analysis to dynamic multilingual interaction [40,41,42].
For instance, in 2020, Weiping Hu’s team at Shaanxi Normal University investigated the elusive neural mechanisms underlying spoken word segmentation in English as a Second Language (ESL) learners. Using fNIRS technology, they designed a word recognition task with two conditions: simple and difficult. They measured the hemodynamic responses in the temporoparietal junction (TPJ) and prefrontal cortex (PFC) to address this issue. The results indicated that the combination of bottom-up sensory input processing (reflected in TPJ activation) and top-down cognitive processing (reflected in PFC activation) is crucial for spoken word segmentation in ESL listeners [43]. This finding about the role of different brain regions in second-language processing provides a basis for further exploring how bilingual experiences affect these regions. In 2021, Borja Blanco’s team measured resting-state functional connectivity in 99 four-month-old monolingual and bilingual infants. They found that the influence of bilingual experience on brain functional connectivity may not be observable under resting-state conditions but might only manifest during explicit linguistic tasks or at later stages of development. This suggests that the impact of bilingualism on brain regions related to language processing may be task-dependent, which is in line with the idea that different language tasks can activate specific brain areas as shown in Hu’s study [44].
In 2024, Neelima Wagley’s team at Arizona State University conducted an in-depth investigation into the specificity of syntactic and semantic processing in children (aged 7 to 11 years) who were exposed to both Spanish and English early in life. The study encompassed 65 participants and revealed significant activation in the left inferior frontal gyrus (IFG) during a morphosyntactic task, whereas the MTG exhibited significant activation during a semantic task. Further task comparison analysis underscored the specialized role of the left superior temporal gyrus (STG) in morphosyntactic processing and the specialized functions of the left MTG and angular gyrus in semantic processing. Notably, although proficiency in either language alone was not uniquely associated with the specialization of any specific brain region, the collective skills in both languages did reflect the involvement of the left MTG in semantic processing and the left IFG in syntactic processing [45]. These results on the specialization of brain regions in bilingual children’s language processing can be related to the cross-linguistic transfer effects. A study led by Jia-Wei Kou at National Taiwan University revealed significant cross-linguistic transfer effects in bilingual children performing Chinese and English tasks. The findings indicate that performance in the second language (L2) influences processing strategies in the native language (L1), particularly on phonological-awareness tasks. At the neural level, it reveals distinctive patterns of brain activation in bilingual children, which may be related to the specialized brain regions involved in different language processes as found by Wagley’s team [46].
In recent years, research into brain activity during communication has deepened, revealing that interactions between specific higher-order brain regions may serve as the neural foundation that facilitates mutual understanding between individuals. Particularly in the context of language communication, studies have found that brain synchrony not only involves low-level language regions but also occurs significantly in higher-order language areas and regions beyond language-specific areas [47,48]. In 2023, Xinyi Jiang’s Master’s thesis examined how native- versus second-language processing alters speaker–listener neural coupling in real-life conversational settings: using fNIRS hyperscanning, the author simultaneously recorded 33 dyads (66 Mandarin–English bilinguals) during a turn-taking listen-speak paradigm comprising three Chinese and three English blocks (90 s per trial). Results revealed significant inter-brain synchrony enhancements in sensorimotor, superior temporal, and prefrontal regions regardless of language. Critically, stronger dorsolateral prefrontal–angular gyrus coupling emerged in the second-language (English) condition, and this coupling correlated with comprehension level, spoken-language proficiency, and gender. This finding about inter-brain synchrony in bilingual communication can be associated with the previous studies on brain regions’ activation in bilingual language processing. The specialized brain regions involved in different language tasks may contribute to the patterns of inter-brain synchrony during communication [49].
In sign language-related cross-linguistic research, Evelyne Mercure’s team employed fNIRS to compare language activation across three groups of children: typically hearing children, deaf children (exposed only to sign language), and hearing children of deaf adults (exposed to both sign and spoken language). The study revealed that bilingual experiences during infancy shape the activation of language networks, demonstrating that unimodal bilingual exposure has a stronger influence on early brain lateralization than bimodal exposure. This work provides important evidence that early language experiences can shape brain development across different modalities [50].

3.5. Clinical Applications

In the clinical field, fNIRS technology can be utilized to evaluate and monitor the treatment progress of patients with language disorders, specifically examining the brain reorganization patterns observed in aphasia patients. By comparing the brain activity of patients with that of healthy controls, clinicians can devise more effective treatment plans. Here we outline fNIRS applications in diagnosing, monitoring, and characterizing cortical reorganization in aphasia, stuttering, and language impairments associated with psychiatric disorders.
In 2020, Lindsay K. Butler’s team at Boston University conducted a comprehensive analysis of 34 fNIRS studies, which involved patients with nine categories of speech or language disorders. The conclusion of this article indicates that, in the realm of speech and language disorders, fNIRS may offer advantages over other neuroimaging techniques. It holds the potential to enhance early and differential diagnosis, improve our understanding of treatment responses, augment neuroprosthetic functionality, and advance neurofeedback interventions [51]. This broad overview of fNIRS’s potential in speech and language disorders sets the stage for more in-depth investigations into specific disorders. For instance, studies on stuttering can build on this general understanding to explore the neural mechanisms unique to this condition.
In 2019, Eric S. Jackson’s team at New York University examined the neural correlates of speech planning and execution in adults who stutter (AWS). Using fNIRS, they measured the neural activity of 15 AWS participants and 15 control participants while they performed specific tasks. The analysis revealed that AWS exhibited underactivation in the speech-language network of the left hemisphere and overactivation in the right hemisphere, findings that were consistent with previous PET/fMRI studies. Additionally, under conditions of high planning load, AWS demonstrated atypical activation in the left hemisphere, whereas under conditions of high execution load, atypical activation was observed in the right frontal, temporal, and parietal regions, as well as in bilateral motor areas. These findings offered new insights into the distinctions in speech planning and execution between AWS and control participants [52]. Building on the understanding of neural differences in AWS, in 2021, the same research group investigated the neural mechanisms underlying speech production in children who stutter (CWS) by implementing tasks designed to manipulate speech planning and execution loads. The study revealed that during speech planning, CWS exhibited atypical activation in the bilateral inferior frontal gyrus and the right supramarginal gyrus. During execution, atypical activation was observed in the bilateral anterior central gyrus, inferior frontal gyrus, right supramarginal gyrus, and superior temporal gyrus. These activation patterns were similar to those previously reported in AWS, but there were also notable differences. Both CWS and AWS showed atypical activation in the right inferior frontal gyrus and supramarginal gyrus during execution. However, only CWS exhibited this pattern during planning. Additionally, during execution, CWS demonstrated atypical activation in the left inferior frontal gyrus and right anterior central gyrus, which was not observed in AWS. These differences between AWS and CWS further enrich our knowledge of stuttering-related neural mechanisms and may help in developing age-specific treatment strategies [53]. Despite its feasibility and numerous advantages, the application of fNIRS in aphasia research, particularly in patients with global aphasia, has been limited.
The Verbal Fluency Task (VFT) is a neuropsychological paradigm that engages multiple cognitive domains and provides an index of how fluently individuals can generate and transmit linguistic information. The combined use of fNIRS and VFT offers direct neurophysiological evidence of executive function and has been widely applied to the diagnosis and differential diagnosis of major depressive disorder, bipolar disorder, and schizophrenia [54,55]. For example, Yeung et al. conducted a systematic review and meta-analysis employing fNIRS during VFT to detect depression, schizophrenia, and other psychiatric conditions, demonstrating that fNIRS measures, particularly increases in HbO, are more sensitive for psychopathological detection under VFT conditions, thereby underscoring the unique value of fNIRS [54]. This application of fNIRS in psychiatric disorders can be related to its use in language-related disorders. Since language impairments are often associated with psychiatric disorders, the neural patterns detected by fNIRS during VFT in psychiatric patients may also provide insights into the language-related neural mechanisms in these conditions.
In 2023, Haozheng Li’s team at the School of Rehabilitation of Shanghai University of Traditional Chinese Medicine examined the cortical activation patterns of global aphasia patients during naming and repetition tasks. The results demonstrated that the left supplementary motor cortex plays a pivotal role in the language functions of global aphasia patients, especially in naming and repetition abilities [56]. In the same year, the team led by Meier EL at Johns Hopkins University utilized fNIRS to investigate the differences in resting-state functional connectivity (rs-FC) between early-stage aphasia patients and healthy adults, as well as its correlation with language impairments. The findings indicated that rs-FC was generally decreased in acute-phase patients, whereas there was no statistically significant difference between subacute-phase patients and the control group. After accounting for the number of days post-stroke, stronger rs-FC within and between hemispheres, particularly in the right hemisphere, was associated with less severe aphasia. Preliminary evidence hints that fNIRS-based rs-FC measurements may serve as an indicator of the early impact of stroke on aphasia patients [57]. The findings of Li’s team on cortical activation patterns and Meier EL’s team on rs-FC in aphasia patients are complementary. Li’s study focuses on the specific cortical regions involved in language tasks, while Meier EL’s study looks at the overall functional connectivity in the brain. Together, they provide a more comprehensive understanding of the neural basis of aphasia and may guide the development of more targeted rehabilitation strategies.

3.6. Applications of Multimodal Neuroimaging Techniques

In recent years, the integration of fNIRS with electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) has emerged as a powerful tool for unveiling the neural dynamics of language. EEG offers millisecond-level temporal precision, fMRI provides millimeter-level spatial resolution, and fNIRS endows the system with high ecological portability. This section compares fNIRS with fMRI and EEG, and illustrates how hybrid approaches (e.g., fNIRS-fMRI, fNIRS-EEG) complement each other to enhance spatial and temporal resolution in language research.
At the cross-modal validation level, the combined fMRI–fNIRS paradigm has been systematically applied to preoperative language function assessment. Arun et al. found moderate consistency between the two modalities in frontal language lateralization [58]. Vannasing et al. reported a case of a child with refractory epilepsy, where fNIRS-EEG and fMRI jointly revealed right-sided language reorganization, highlighting the ecological supplementary value of fNIRS to fMRI results [59]. Tung et al. captured network reorganization during a verbal fluency task in patients with fronto-temporal epilepsy using fNIRS, which was independently validated by fMRI [60]. Meier et al. (2023) further cross-validated resting-state fNIRS functional connectivity with fMRI in patients with acute and subacute aphasia, confirming cross-modal test-retest reliability [57].
In recent years, the application of EEG–fNIRS bimodal technology in language and cognitive research has continued to expand, forming a complete evidence chain from “algorithm innovation–developmental mechanisms–clinical translation”. First, at the algorithmic level, Cooney, C et al. proposed a novel EEG–fNIRS bimodal deep learning architecture for decoding overt and imagined speech [61]. After 19 participants completed cue-guided speech tasks, a hybrid model incorporating convolutional subnetworks enhanced the classification accuracy of overt/imagined speech from single-modality levels to 87% and 53%, respectively, demonstrating that multimodal deep learning can significantly enhance neural signal interpretation capabilities. Secondly, at the level of developmental mechanisms, two fNIRS studies have separately elucidated the phonetic processing patterns in newborns and bilingual children: ① Combining EEG-fNIRS, Cabrera, L. et al. demonstrated that newborns, just hours old, could distinguish/pa/-/ta/using amplitude modulation (AM) and frequency modulation (FM) temporal cues [62]. Fast modulation elicited stronger fronto-temporo-parietal hemodynamic responses, suggesting that ultra-early phoneme encoding already possesses adult-like mechanisms; weakened activation under slow modulation revealed limitations in processing complex acoustic information. These findings provide actionable multimodal markers for optimizing early language input and tracking auditory-language development. ② Research on Chinese-English bilingual children revealed that performance on L2 English tasks could cross-linguistically enhance L1 Chinese phonological awareness, presenting unique activation patterns in regions such as the left dorsolateral prefrontal cortex (DLPFC). This study, for the first time, delineated the cross-linguistic transfer effects of bilingual education at the neural level [46]. Finally, at the clinical translation level, an EEG–fNIRS study conducted by the Beijing Institute of Otolaryngology included 78 elderly participants (26 each in the normal hearing/mild hearing impairment/moderate-severe hearing impairment groups). It found that with age-related hearing loss aggravation, the left DLPFC exhibited decompensation, with reduced theta-band connectivity impairing top-down speech prediction. The study further suggested that the DLPFC could serve as an intervention target to improve speech recognition in noisy environments [63].
Synthesizing the above content, we constructed a timeline diagram as shown in Figure 2. The horizontal arrows, extending from infancy to adulthood, represent the natural developmental progression of human language abilities from basic phonemic perception to complex language processing. The vertical arrows, on the other hand, illustrate the expansion of different fNIRS research themes within the same developmental stage, reflecting how the application of the technology has evolved in response to the deepening demands of language development research. The references cited in the figure represent foundational or validating studies in their respective research directions.

4. Limitations of fNIRS

fNIRS offers unique advantages in neuroimaging, yet several inherent limitations must be acknowledged.
(1)
Limited spatial resolution and penetration depth: fNIRS predominantly samples cortical surface hemodynamics, with an effective penetration of ~1.5–2 cm, thereby providing restricted sensitivity to subcortical structures such as the thalamus or brainstem [64]. Recent efforts to mitigate this constraint include the incorporation of individual MRI-based structural templates [65], smartphone-based photogrammetry [66], and high-density diffuse optical tomography systems [67].
(2)
Signal contamination: Hemodynamic fluctuations originating from extracerebral tissues—scalp blood flow, skull thickness, and hair pigmentation—can introduce significant artifacts, potentially compromising signal fidelity [68]. Emerging hardware developments are actively optimizing optical coupling and wavelength selection to attenuate these confounds, thereby reducing participant exclusion attributable to skin or hair characteristics [69,70,71]. Current best-practice combines: (i) individualized 3-D printed helmets with real-time pressure feedback [72]; (ii) refractive-index-matching gel or transparent patches [73]; and (iii) flexible silicone arrays optimized for age-specific head geometry together with post hoc weighted-regression algorithms [74].
(3)
Lack of standardized pipelines: The absence of universally adopted preprocessing and statistical procedures hampers cross-study comparability and reproducibility [73,75].
Table 2 summarizes the comparison and concordance between fNIRS and other brain imaging techniques. The complementary strengths of these three modalities enable researchers to simultaneously pinpoint the “when” and “where” of language-related activities, thereby obtaining a more comprehensive picture of brain language processing. Despite these limitations, fNIRS remains a valuable, non-invasive, and portable modality for real-time brain-function assessment in both clinical and ecological settings. Continued technological refinement and multimodal integration are expected to further enhance its performance and broaden its applicability.

5. Critical Assessment

Regarding the application of fNIRS in speech perception and recognition, multiple independent studies (such as those conducted by Ming Dong’s team and Jessica Defenderfer’s team) have demonstrated that fNIRS can reliably detect cortical activation during speech processing, particularly in classic language-related regions such as Broca’s area and Wernicke’s area. These findings have been consistently replicated across different laboratories and participant groups, indicating their high reliability and validity. However, for some newer research areas, such as the application of fNIRS in cross-linguistic processing and bilingual studies, the current results are mostly preliminary observations.
From the perspective of the overall research in the field, multiple teams consistently support the core conclusion that “bilingual processing relies on the left-hemisphere language network.” For instance, both the Wagley team and the Kou team identified the critical roles of regions such as the left inferior frontal gyrus and the left middle temporal gyrus [45,46]. However, inconsistencies exist regarding the “neural localization of cross-language transfer effects.” The Kou team associated activation in the left inferior frontal gyrus with L2-to-L1 phonological transfer, whereas the Blanco team, studying Spanish-English bilingual children aged 7–9, found that activation in the left middle frontal gyrus better reflected bilingual phonological transfer effects [74]. The Sebastian team suggested that differences in the neural localization of cross-language transfer effects may be related to language typology (tonal vs. non-tonal languages) and the age of bilingual acquisition (early vs. late) [76].

6. Future Perspectives

Looking forward, the application of fNIRS in linguistics is poised to expand and evolve in several significant ways. As summarized in Table 3, which outlines key language-related brain regions and their functions as revealed by fNIRS, we can see the specific areas of the brain involved in various linguistic processes. This foundational knowledge sets the stage for multiple promising directions.
One particularly promising direction is the advancement of fNIRS technology itself. The enhancement of fNIRS spatial resolution is primarily achieved through three approaches: First, optimizing probe design using tools such as fNIRS Optodes’ Location Designer, which determines potential probe positions based on brain atlases and standard EEG locations [77]. For more complex layouts, tools like Array Designer can generate sensitivity profiles by setting parameters (e.g., number and distance of source-detector pairs) to optimize probe design [78]. Second, precise coregistration techniques align individual MRI anatomical images with 3D optode positions to map channels onto the cortical surface. The team led by Ippeita Dan developed a transformation algorithm based on the international 10-20/10-5 EEG system, converting scalp optode locations to standard brain space and addressing the core challenge of lacking anatomical reference in fNIRS localization [79]. Brigadoi, Lloyd-Fox, and others developed tools such as AtlasViewer (https://github.com/BUNPC/AtlasViewer (accessed on 1 November 2025)) and Nirstorm (https://github.com/Nirstorm/nirstorm (accessed on 1 November 2025)), integrated into the Brainstorm software, which support fMRI-fNIRS coregistration and layered head modeling, significantly improving channel localization accuracy [78,80]. Third, accurate headgear placement, a common method derived from electroencephalography, involves standardized positioning of the headset according to established systems such as the international 10–20, 10–10, or 10–5 EEG systems [81,82]. This enhanced capability will facilitate more detailed studies of language-related brain regions, thereby providing deeper insights into how specific areas contribute to various linguistic functions.
Another exciting development is the integration of fNIRS with other neuroimaging modalities, such as EEG and fMRI. By combining the temporal resolution of EEG, the spatial resolution of fMRI, and the portability of fNIRS, researchers can gain a more comprehensive understanding of the dynamics of language processing. This approach allows them to capture both the timing and the location of language-related brain activity, providing a richer picture of how the brain processes language.
In clinical settings, fNIRS is anticipated to become an increasingly widely used tool for diagnosing and monitoring language disorders. Its non-invasive nature, coupled with its suitability for use with a diverse range of populations, including infants and individuals with mobility limitations, renders it ideally suited for conducting regular assessments of language function over time. As fNIRS technology becomes more affordable and accessible, it has the potential to become a standard component of the diagnostic toolkit for speech-language pathologists and neurologists.
The literature cited in this study is primarily sourced from Chinese and English databases, which to some extent limits the generalizability and universality of the research conclusions to other linguistic contexts. Significant differences exist among different languages in terms of phonetics, vocabulary, syntax, and pragmatics, and these differences may affect the application effectiveness of fNIRS in language processing research. Future studies should further expand multilingual samples to validate the effectiveness and reliability of fNIRS technology across diverse linguistic backgrounds.
Finally, the growing use of fNIRS in multilingual and cross-cultural studies holds the potential to deepen our understanding of how different languages are processed in the brain. This research has the capacity to inform the development of language education strategies that are tailored to the unique neural profiles of learners, particularly in multilingual societies. Overall, fNIRS is poised to play a pivotal role in the future of linguistic research, presenting new opportunities to explore the intricate neural underpinnings of language.

7. Conclusions

In conclusion, fNIRS technology holds vast prospects for application in language learning. It not only aids in understanding brain activity during speech perception, language processing, and language acquisition but also furnishes invaluable information for clinical diagnosis and treatment. However, fNIRS technology is confronted with several challenges, including signal interference, intricate data analysis, and limited spatial resolution. Future research must focus on addressing these challenges to enhance the reliability and precision of fNIRS, thereby bolstering its significance in future language acquisition research. Additionally, integrating fNIRS with other brain imaging technologies and behavioral methods could offer a more holistic understanding of the neural mechanisms underlying language acquisition. Future research should pursue continuous technological innovation, integrate multimodal approaches, deepen clinical applications, and expand cross-linguistic investigations to advance the field of language neuroscience. It is anticipated that in the future, fNIRS technology will play an even more pivotal role in the field of language learning.

Author Contributions

Conceptualization, X.Z. (Xiu Zhang) and P.C.; methodology, P.C. and X.Z. (Xin Zhang); validation, P.C. and Y.C.; writing—original draft preparation, P.C.; writing—review and editing, Y.C. and X.Z. (Xiu Zhang); supervision, X.Z. (Xiu Zhang); project administration, X.Z. (Xiu Zhang); funding acquisition, X.Z. (Xiu Zhang). All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Natural Science Foundation of China grant number 61905176.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Piper, S.K.; Krueger, A.; Koch, S.P.; Mehnert, J.; Habermehl, C.; Steinbrink, J.; Obrig, H.; Schmitz, C.H. A wearable multi-channel fNIRS system for brain imaging in freely moving subjects. Neuroimage 2014, 85, 64–71. [Google Scholar] [CrossRef] [PubMed]
  2. McKendrick, R.; Parasuraman, R.; Ayaz, H. Wearable functional near infrared spectroscopy (fNIRS) and transcranial direct current stimulation (tDCS): Expanding vistas for neurocognitive augmentation. Front. Syst. Neurosci. 2015, 9, 27. [Google Scholar] [CrossRef] [PubMed]
  3. Ayaz, H.; Onaral, B.; Izzetoglu, K.; Shewokis, P.A.; McKendrick, R.; Parasuraman, R. Continuous monitoring of brain dynamics with functional near infrared spectroscopy as a tool for neuroergonomic research: Empirical examples and a technological development. Front. Hum. Neurosci. 2013, 7, 871. [Google Scholar] [CrossRef] [PubMed]
  4. Liu, Y.; Ayaz, H.; Shewokis, P.A. Multisubject “Learning” for Mental Workload Classification Using Concurrent EEG, fNIRS, and Physiological Measures. Front. Hum. Neurosci. 2017, 11, 389. [Google Scholar] [CrossRef]
  5. Beauchamp, M.S.; Beurlot, M.R.; Fava, E.; Nath, A.R.; Parikh, N.A.; Saad, Z.S.; Bortfeld, H.; Oghalai, J.S. The developmental trajectory of brain-scalp distance from birth through childhood: Implications for functional neuroimaging. PLoS ONE 2011, 6, e24981. [Google Scholar] [CrossRef]
  6. Kabdebon, C.; Leroy, F.; Simmonet, H.; Perrot, M.; Dubois, J.; Dehaene-Lambertz, G. Anatomical correlations of the international 10-20 sensor placement system in infants. Neuroimage 2014, 99, 342–356. [Google Scholar] [CrossRef]
  7. Calderon-Arnulphi, M.; Alaraj, A.; Slavin, K.V. Near infrared technology in neuroscience: Past, present and future. Neurol. Res. 2009, 31, 605–614. [Google Scholar] [CrossRef]
  8. Chance, B.; Zhuang, Z.; UnAh, C.; Alter, C.; Lipton, L. Cognition-activated low-frequency modulation of light absorption in human brain. Proc. Natl. Acad. Sci. USA 1993, 90, 3770–3774. [Google Scholar] [CrossRef]
  9. Villringer, A.; Planck, J.; Hock, C.; Schleinkofer, L.; Dirnagl, U. Near infrared spectroscopy (NIRS): A new tool to study hemodynamic changes during activation of brain function in human adults. Neurosci. Lett. 1993, 154, 101–104. [Google Scholar] [CrossRef]
  10. Gervain, J.; Mehler, J.; Werker, J.F.; Nelson, C.A.; Csibra, G.; Lloyd-Fox, S.; Shukla, M.; Aslin, R.N. Near-infrared spectroscopy: A report from the McDonnell infant methodology consortium. Dev. Cogn. Neurosci. 2011, 1, 22–46. [Google Scholar] [CrossRef]
  11. Lloyd-Fox, S.; Blasi, A.; Elwell, C.E. Illuminating the developing brain: The past, present and future of functional near infrared spectroscopy. Neurosci. Biobehav. Rev. 2010, 34, 269–284. [Google Scholar] [CrossRef] [PubMed]
  12. Pinti, P.; Tachtsidis, I.; Hamilton, A.; Hirsch, J.; Aichelburg, C.; Gilbert, S.; Burgess, P.W. The present and future use of functional near-infrared spectroscopy (fNIRS) for cognitive neuroscience. Ann. N. Y. Acad. Sci. 2020, 1464, 5–29. [Google Scholar] [CrossRef] [PubMed]
  13. Gallagher, A.; Béland, R.; Lassonde, M. The contribution of functional near-infrared spectroscopy (fNIRS) to the presurgical assessment of language function in children. Brain Lang. 2012, 121, 124–129. [Google Scholar] [CrossRef] [PubMed]
  14. Morrel, J.; Singapuri, K.; Landa, R.J.; Reetzke, R. Neural correlates and predictors of speech and language development in infants at elevated likelihood for autism: A systematic review. Front. Hum. Neurosci. 2023, 17, 1211676. [Google Scholar] [CrossRef]
  15. Shen, Q.Q.; Hou, J.M.; Xia, T.; Zhang, J.Y.; Wang, D.L.; Yang, Y.; Luo, R.; Xin, Z.L.; Yin, H.C.; Cui, L. Exercise promotes brain health: A systematic review of fNIRS studies. Front. Psychol. 2024, 15, 1327822. [Google Scholar] [CrossRef]
  16. Jöbsis, F.F. Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977, 198, 1264–1267. [Google Scholar] [CrossRef]
  17. Strangman, G.; Culver, J.P.; Thompson, J.H.; Boas, D.A. A quantitative comparison of simultaneous BOLD fMRI and NIRS recordings during functional brain activation. Neuroimage 2002, 17, 719–731. [Google Scholar] [CrossRef]
  18. Matcher, S.J.; Kirkpatrick, P.; Nahid, K.; Cope, M.; Delpy, D.T. Absolute quantification methods in tissue near-infrared spectroscopy. Proc. SPIE 1995, 2389, 486–495. [Google Scholar]
  19. Cope, M.; Delpy, D.T. System for long-term measurement of cerebral blood and tissue oxygenation on newborn infants by near infra-red transillumination. Med. Biol. Eng. Comput. 1988, 26, 289–294. [Google Scholar] [CrossRef]
  20. Scholkmann, F.; Wolf, M. General equation for the differential pathlength factor of the frontal human head depending on wavelength and age. J. Biomed. Opt. 2013, 18, 105004. [Google Scholar] [CrossRef]
  21. Torricelli, A.; Contini, D.; Pifferi, A.; Arpaia, L.; Spinelli, L.; Cubeddu, R.; Martelli, F. Time domain functional NIRS imaging for human brain mapping. NeuroImage 2014, 85, 28–50. [Google Scholar] [CrossRef] [PubMed]
  22. Ding, P.; Zhou, Z.; Geng, H.; Zhang, K.; Wang, H.; Li, Y.; Li, X. A cost-effective fNIRS system for multisubject multibrain regions hyperscanning. IEEE Trans. Instrum. Meas. 2025, 74, 4505621. [Google Scholar] [CrossRef]
  23. Zhou, X.; Xia, Y.; Uchitel, J.; Collins-Jones, L.; Yang, S.; Loureiro, R.; Cooper, R.J.; Zhao, H. Review of recent advances in frequency-domain near-infrared spectroscopy technologies [Invited]. Biomed. Opt. Express 2023, 14, 3234–3258. [Google Scholar] [CrossRef] [PubMed]
  24. Si, X.; Li, S.; Xiang, S.; Yu, J.; Ming, D. Imagined speech increases the hemodynamic response and functional connectivity of the dorsal motor cortex. J. Neural Eng. 2021, 18, 056048. [Google Scholar] [CrossRef]
  25. Butera, I.M.; Larson, E.D.; DeFreese, A.J.; Lee, A.K.; Gifford, R.H.; Wallace, M.T. Functional localization of audiovisual speech using near infrared spectroscopy. Brain Topogr. 2022, 35, 416–430. [Google Scholar] [CrossRef]
  26. Defenderfer, J.; Forbes, S.; Wijeakumar, S.; Hedrick, M.; Plyler, P.; Buss, A.T. Frontotemporal activation differs between perception of simulated cochlear implant speech and speech in background noise: An image-based fNIRS study. Neuroimage 2021, 240, 118385. [Google Scholar] [CrossRef]
  27. Wu, Y.J.; Hou, X.; Peng, C.; Yu, W.; Oppenheim, G.M.; Thierry, G.; Zhang, D. Rapid learning of a phonemic discrimination in the first hours of life. Nat. Hum. Behav. 2022, 6, 1169–1179. [Google Scholar] [CrossRef]
  28. Zhang, H.; Guo, Z.; Chen, F. The Effects of Different Brain Regions on fNIRS-based Task-state Detection in Speech Imagery. In Proceedings of the 2023 45th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), Sydney, Australia, 24–27 July 2023; pp. 1–4. [Google Scholar]
  29. Zhang, F.; Gervain, J.; Roeyers, H. Developmental changes in the brain response to speech during the first year of life: A near-infrared spectroscopy study of dutch-learning infants. Infant Behav. Dev. 2022, 67, 101724. [Google Scholar] [CrossRef]
  30. Jia, G.; Liu, G.; Niu, H. Hemispheric Lateralization of Visuospatial Attention Is Independent of Language Production on Right-Handers: Evidence from Functional Near-Infrared Spectroscopy. Front. Neurol. 2022, 12, 784821. [Google Scholar] [CrossRef]
  31. Wilcox, T.; Biondi, M. fNIRS in the developmental sciences. Wiley Interdiscip. Rev. Cogn. Sci. 2015, 6, 263–283. [Google Scholar] [CrossRef]
  32. Hoff, E.; Rumiche, R.L. Research Methods in Child Language: A Practical Guide; University of Cambodia Library: Phnom Penh, Cambodia, 2012. [Google Scholar]
  33. Altvater-Mackensen, N.; Grossmann, T. Modality-independent recruitment of inferior frontal cortex during speech processing in human infants. Dev. Cogn. Neurosci. 2018, 34, 130–138. [Google Scholar] [CrossRef] [PubMed]
  34. Wen, X.; Zhang, H.; Li, G.; Liu, M.; Yin, W.; Lin, W.; Zhang, J.; Shen, D. First-year development of modules and hubs in infant brain functional networks. Neuroimage 2019, 185, 222–235. [Google Scholar] [CrossRef] [PubMed]
  35. Lawrence, R.J.; Wiggins, I.M.; Hodgson, J.C.; Hartley, D.E.H. Evaluating cortical responses to speech in children: A functional near-infrared spectroscopy (fNIRS) study. Hear. Res. 2021, 401, 108155. [Google Scholar] [CrossRef] [PubMed]
  36. Yu, J. The Role and Neural Mechanisms of Consonant, Vowel and Tone in Chinese Language Output: An fNIRS Study. Master’s Thesis, Northeast Normal University, Changchun, China, 2023. [Google Scholar]
  37. Martinez-Alvarez, A.; Gervain, J.; Koulaguina, E.; Pons, F.; de Diego-Balaguer, R. Prosodic cues enhance infants’ sensitivity to nonadjacent regularities. Sci. Adv. 2023, 9, eade4083. [Google Scholar] [CrossRef]
  38. Ren, J.; Cai, L.; Jia, G.; Niu, H. Cortical specialization associated with native speech category acquisition in early infancy. Cereb. Cortex 2024, 34, 124. [Google Scholar] [CrossRef]
  39. Zhou, X.; Wang, L.; Hong, X.; Wong, P.C. Infant-directed speech facilitates word learning through attentional mechanisms: An fNIRS study of toddlers. Dev. Sci. 2024, 27, e13424. [Google Scholar]
  40. Liu, Y.; Piazza, E.A.; Simony, E.; Shewokis, P.A.; Onaral, B.; Hasson, U.; Ayaz, H. Measuring speaker-listener neural coupling with functional near infrared spectroscopy. Sci Rep. 2017, 7, 43293. [Google Scholar]
  41. Hull, R.; Bortfeld, H.; Koons, S. Near-Infrared Spectroscopy and Cortical Responses to Speech Production. Open Neuroimaging J. 2009, 3, 26–30. [Google Scholar] [CrossRef]
  42. Papoutselou, E. Exploring Neural Markers of Language Processing Using fNIRS. Ph.D. Thesis, University of Nottingham, Nottingham, UK, 2023. [Google Scholar]
  43. Li, Y.; Yang, Y.; Tang, A.C.; Liu, N.; Wang, X.; Du, Y.; Hu, W. English spoken word segmentation activates the prefrontal cortex and temporo-parietal junction in Chinese ESL learners: A functional near-infrared spectroscopy (fNIRS) study. Brain Res. 2020, 1733, 146693. [Google Scholar]
  44. Blanco, B.; Molnar, M.; Carreiras, M.; Collins-Jones, L.H.; Vidal, E.; Cooper, R.J.; Caballero-Gaudes, C. Group-level cortical functional connectivity patterns using fNIRS: Assessing the effect of bilingualism in young infants. Neurophotonics 2021, 8, 025011. [Google Scholar]
  45. Wagley, N.; Hu, X.; Satterfield, T.; Bedore, L.M.; Booth, R.J.; Kovelman, I. Neural specificity for semantic and syntactic processing in Spanish-English bilingual children. Brain Lang. 2024, 250, 105380. [Google Scholar] [CrossRef] [PubMed]
  46. Kou, J.W.; Fan, L.Y.; Chen, H.C.; Chen, S.Y.; Hu, X.; Zhang, K.; Kovelman, I.; Chou, T.L. Neural substrates of L2-L1 transfer effects on phonological awareness in young Chinese-English bilingual children. NeuroImage 2024, 291, 120592. [Google Scholar] [CrossRef] [PubMed]
  47. Stephens, G.J.; Silbert, L.J.; Hasson, U. Speaker-listener neural coupling underlies successful communication. Proc. Natl. Acad. Sci. USA 2010, 107, 14425–14430. [Google Scholar] [CrossRef] [PubMed]
  48. Liu, L.; Zhang, Y.; Zhou, Q.; Garrett, D.D.; Lu, C.; Chen, A.; Qiu, J.; Ding, G. Auditory-Articulatory Neural Alignment between Listener and Speaker during Verbal Communication. Cereb. Cortex 2020, 30, 942–951. [Google Scholar] [CrossRef]
  49. Jiang, X. Brain Synchrony of Chinese-English Bilinguals During First or Second Language Listening and Speaking: An fNIRS Hyperscanning Study. Master’s Thesis, East China Normal University, Shanghai, China, 2023. [Google Scholar]
  50. Mercure, E.; Evans, S.; Pirazzoli, L.; Goldberg, L.; Bowden-Howl, H.; Coulson-Thaker, K.; Beedie, I.; Lloyd-Fox, S.; Johnson, M.H.; MacSweeney, M. Language Experience Impacts Brain Activation for Spoken and Signed Language in Infancy: Insights from Unimodal and Bimodal Bilinguals. Neurobiol. Lang. 2020, 1, 9–32. [Google Scholar] [CrossRef]
  51. Butler, L.K.; Kiran, S.; Tager-Flusberg, H. Functional Near-Infrared Spectroscopy in the Study of Speech and Language Impairment Across the Life Span: A Systematic Review. Am. J. Speech Lang. Pathol. 2020, 29, 1674–1701. [Google Scholar]
  52. Jackson, E.S.; Wijeakumar, S.; Beal, D.S.; Brown, B.; Zebrowski, P.; Spencer, J.P. A fNIRS Investigation of Speech Planning and Execution in Adults Who Stutter. Neuroscience 2019, 406, 73–85. [Google Scholar] [CrossRef]
  53. Jackson, E.S.; Wijeakumar, S.; Beal, D.S.; Brown, B.; Zebrowski, P.M.; Spencer, J.P. Speech planning and execution in children who stutter: Preliminary findings from a fNIRS investigation. J. Clin. Neurosci. 2021, 91, 32–42. [Google Scholar] [CrossRef]
  54. Yeung, M.K.; Lin, J. Probing depression, schizophrenia, and other psychiatric disorders using fNIRS and the verbal fluency test: A systematic review and meta-analysis. J. Psychiatr. Res. 2021, 140, 416–435. [Google Scholar] [CrossRef]
  55. Husain, F.S.; Tang, T.; Yu, R.; Tam, W.W.; Tran, B.; Quek, T.T.; Hwang, S.H.; Chang, C.W.; Ho, C.S.; Ho, R.C. Cortical haemodynamic response measured by functional near infrared spectroscopy during a verbal fluency task in patients with major depression and borderline personality disorder. EBioMedicine 2020, 51, 102586. [Google Scholar] [CrossRef]
  56. Li, H.; Liu, J.; Tian, S.; Fan, S.; Wang, T.; Qian, H.; Liu, G.; Zhu, Y.; Wu, Y.; Hu, R. Language reorganization patterns in global aphasia-evidence from fNIRS. Front. Neurol. 2023, 13, 1025384. [Google Scholar] [CrossRef] [PubMed]
  57. Meier, E.L.; Bunker, L.D.; Kim, H.; Hillis, A.E. Resting-State Connectivity in Acute and Subacute Poststroke Aphasia: A Functional Near-Infrared Spectroscopy Pilot Study. Brain Connect. 2023, 13, 441–452. [Google Scholar] [CrossRef] [PubMed]
  58. Arun, K.; Smitha, K.; Rajesh, P.; Kesavadas, C. Functional near-infrared spectroscopy is in moderate accordance with functional MRI in determining lateralisation of frontal language areas. Neuroradiol. J. 2017, 31, 197140091773908. [Google Scholar] [CrossRef] [PubMed]
  59. Vannasing, P.; Cornaggia, I.; Vanasse, C.; Tremblay, J.; Diadori, P.; Perreault, S.; Lassonde, M.; Gallagher, A. Potential brain language reorganization in a boy with refractory epilepsy; an fNIRS–EEG and fMRI comparison. Epilepsy Behav. Case Rep. 2016, 5, 34–37. [Google Scholar] [CrossRef]
  60. Tung, H.; Lin, W.H.; Lan, T.H.; Hsieh, P.F.; Chiang, M.C.; Lin, Y.Y.; Peng, S.J. Network reorganization during verbal fluency task in fronto-temporal epilepsy: A functional near-infrared spectroscopy study. J. Psychiatr. Res. 2021, 138, 541–549. [Google Scholar] [CrossRef]
  61. Cooney, C.; Folli, R.; Coyle, D. A bimodal deep learning architecture for EEG-fNIRS decoding of overt and imagined speech. IEEE Trans. Biomed. Eng. 2022, 69, 1983–1993. [Google Scholar] [CrossRef]
  62. Cabrera, L.; Gervain, J. Speech perception at birth: The brain encodes fast and slow temporal information. Sci Adv. 2020, 6, eaba7830. [Google Scholar] [CrossRef]
  63. Wang, S.; Liu, Y.; Kou, N.; Chen, Y.; Liu, T.; Wang, Y.; Wang, S. Impact of age-related hearing loss on decompensation of left DLPFC during speech perception in noise: A combined EEG-fNIRS study. GeroScience 2025, 47, 2119–2134. [Google Scholar] [CrossRef]
  64. Cui, X.; Bray, S.; Bryant, D.M.; Glover, G.H.; Reiss, A.L. A quantitative comparison of NIRS and fMRI across multiple cognitive tasks. NeuroImage 2011, 54, 2808–2821. [Google Scholar] [CrossRef]
  65. Aasted, C.M.; Yücel, M.A.; Cooper, R.J.; Dubb, J.; Tsuzuki, D.; Becerra, L.; Petkov, M.P.; Borsook, D.; Dan, I.; Boas, D.A. Anatomical guidance for functional nearinfrared spectroscopy: AtlasViewer tutorial. Neurophotonics 2025, 2, 020801. [Google Scholar] [CrossRef]
  66. Mazzonetto, I.; Castellaro, M.; Cooper, R.J.; Brigadoi, S. Smartphone-based photogrammetry provides improved localization and registration of scalp-mounted neuroimaging sensors. Sci. Rep. 2022, 12, 10862. [Google Scholar] [CrossRef]
  67. Eggebrecht, A.T.; Ferradal, S.L.; Robichaux-Viehoever, A.; Hassanpour, M.S.; Dehghani, H.; Snyder, A.Z.; Hershey, T.; Culver, J.P. Mapping distributed brain function and networks with diffuse optical tomography. Nat. Photonics 2014, 8, 448–454. [Google Scholar] [CrossRef]
  68. Kwasa, J.; Peterson, H.M.; Karrobi, K.; Jones, L.; Parker, T.; Nickerson, N.; Wood, S. Demographic reporting and phenotypic exclusion in fNIRS. Front. Neurosci. 2023, 17, 1086208. [Google Scholar] [CrossRef]
  69. Doherty, E.J.; Spencer, C.A.; Burnison, J.; Čeko, M.; Chin, J.; Eloy, L.; Haring, K.; Kim, P.; Pittman, D.; Powers, S.; et al. Interdisciplinary views of fNIRS: Current advancements, equity challenges, and an agenda for future needs of a diverse fNIRS research community. Front. Integr. Neurosci. 2023, 17, 1059679. [Google Scholar] [CrossRef]
  70. Perdue, K.L.; Jensen, S.K.G.; Kumar, S.; Richards, J.E.; Kakon, S.H.; Haque, R.; Petri, W.A., Jr.; Lloyd-Fox, S.; Elwell, C.; Nelson, C.A. Using functional near-infrared spectroscopy to assess social information processing in poor urban Bangladeshi infants and toddlers. Dev. Sci. 2019, 22, e12839. [Google Scholar] [PubMed]
  71. Wijeakumar, S.; Kumar, A.; Delgado Reyes, L.M.; Tiwari, M.; Spencer, J.P. Early adversity in rural India impacts the brain networks underlying visual working memory. Dev. Sci. 2019, 22, e12822. [Google Scholar] [CrossRef] [PubMed]
  72. Collins-Jones, L.H.; Arichi, T.; Poppe, T.; Billing, A.; Xiao, J.; Fabrizi, L.; Brigadoi, S.; Hebden, J.C.; Elwell, C.E.; Cooper, R.J. Construction and validation of a database of head models for functional imaging of the neonatal brain. Hum. Brain Mapp. 2021, 42, 567–586. [Google Scholar] [PubMed]
  73. Yücel, M.A.; Lühmann, A.V.; Scholkmann, F.; Gervain, J.; Dan, I.; Ayaz, H.; Boas, D.; Cooper, R.J.; Culver, J.; Elwell, C.E.; et al. Best practices for fNIRS publications. Neurophotonics 2021, 8, 012101. [Google Scholar]
  74. Frijia, E.M.; Billing, A.; Lloyd-Fox, S.; Vidal Rosas, E.; Collins-Jones, L.; Crespo-Llado, M.M.; Amadó, M.P.; Austin, T.; Edwards, A.; Dunne, L.; et al. Functional imaging of the developing brain with wearable high-density diffuse optical tomography: A new benchmark for infant neuroimaging outside the scanner environment. Neuroimage 2021, 225, 117490. [Google Scholar] [CrossRef]
  75. Di Lorenzo, R.; Pirazzoli, L.; Blasi, A.; Bulgarelli, C.; Hakuno, Y.; Minagawa, Y.; Brigadoi, S. Recommendations for motion correction of infant fNIRS data applicable to multiple data sets and acquisition systems. NeuroImage 2019, 200, 511–527. [Google Scholar] [CrossRef]
  76. Costa, A.; Sebastián-Gallés, N. How does the bilingual experience sculpt the brain? Nat. Rev. Neurosci. 2014, 15, 336–345. [Google Scholar] [CrossRef]
  77. Zimeo Morais, G.A.; Balardin, J.B.; Sato, J.R. fNIRS Optodes’ Location Decider (fOLD): A toolbox for probe arrangement guided by brain regions-of-interest. Sci. Rep. 2018, 8, 3341. [Google Scholar] [CrossRef]
  78. Brigadoi, S.; Salvagnin, D.; Fischetti, M.; Cooper, R.J. Array Designer: Automated optimized array design for functional near-infrared spectroscopy. Neurophotonics 2018, 5, 035010. [Google Scholar] [CrossRef]
  79. Tsuzuki, D.; Dan, I. Spatial registration for functional near-infrared spectroscopy: From channel position on the scalp to cortical location in individual and group analyses. NeuroImage 2014, 85, 92–103. [Google Scholar] [CrossRef]
  80. Delaire, É.; Vincent, T.; Cai, Z.; Machado, A.; Hugueville, L.; Schwartz, D.; Tadel, F.; Cassani, R.; Bherer, L.; Lina, J.-M.; et al. NIRSTORM: A Brainstorm extension dedicated to functional near-infrared spectroscopy data analysis, advanced 3D reconstructions, and optimal probe design. Neurophotonics 2025, 12, 025011. [Google Scholar] [CrossRef]
  81. Klem, G.H.; Lüders, H.O.; Jasper, H.H.; Elger, C. The ten-twenty electrode system of the International Federation. The International Federation of Clinical Neurophysiology. Electroencephalogr. Clin. Neurophysiol. Suppl. 1999, 52, 3–6. [Google Scholar]
  82. Oostenveld, R.; Praamstra, P. The five percent electrode system for high-resolution EEG and ERP measurements. Clin. Neurophysiol. 2001, 112, 713–719. [Google Scholar] [CrossRef]
Photonics 13 00054 g001
Figure 1. Diagram of brain regions associated with speech production and perception.
Figure 1. Diagram of brain regions associated with speech production and perception.
Photonics 13 00054 g001
Photonics 13 00054 g002
Figure 2. Language Development Timeline and fNlRS Research Applications. The horizontal axis is divided into four age periods: infancy, toddlerhood, childhood, and adulthood, marked by gray rounded rectangles that deepen in shade with increasing age. Below each period, colored rectangular blocks indicate the research domains in which fNIRS has been employed. Cope, M. 1988: Reference details for [19]; Si, X. 2023: Reference details for [24]; Butera, I.M. 2022: Reference details for [25]; Defenderfer, J. 2021: Reference details for [26]; Zhang, F. 2022: Reference details for [29]; Altvater-Mackensen, N. 2018: Reference details for [33]; Wen, X. 2019: Reference details for [34]; Lawrence, R.J. 2021: Reference details for [35]; Yu, J. 2023: Reference details for [36]; Martinez-Alvarez, A. 2023: Reference details for [37]; Ren, J. 2024: Reference details for [38]; Zhou, X. 2024: Reference details for [39];Wagley, N. 2024: Reference details for [44]; Wagley, N. 2024: Reference details for [45]; Kou, J.W. 2024: Reference details for [46]; Liu, L. 2020: Reference details for [48]; Mercure, E. 2020: Reference details for [50]; Jackson, E.S. 2021: Reference details for [53]; Li, H. 2023: Reference details for [56]; Meier, E.L. 2023: Reference details for [57].
Figure 2. Language Development Timeline and fNlRS Research Applications. The horizontal axis is divided into four age periods: infancy, toddlerhood, childhood, and adulthood, marked by gray rounded rectangles that deepen in shade with increasing age. Below each period, colored rectangular blocks indicate the research domains in which fNIRS has been employed. Cope, M. 1988: Reference details for [19]; Si, X. 2023: Reference details for [24]; Butera, I.M. 2022: Reference details for [25]; Defenderfer, J. 2021: Reference details for [26]; Zhang, F. 2022: Reference details for [29]; Altvater-Mackensen, N. 2018: Reference details for [33]; Wen, X. 2019: Reference details for [34]; Lawrence, R.J. 2021: Reference details for [35]; Yu, J. 2023: Reference details for [36]; Martinez-Alvarez, A. 2023: Reference details for [37]; Ren, J. 2024: Reference details for [38]; Zhou, X. 2024: Reference details for [39];Wagley, N. 2024: Reference details for [44]; Wagley, N. 2024: Reference details for [45]; Kou, J.W. 2024: Reference details for [46]; Liu, L. 2020: Reference details for [48]; Mercure, E. 2020: Reference details for [50]; Jackson, E.S. 2021: Reference details for [53]; Li, H. 2023: Reference details for [56]; Meier, E.L. 2023: Reference details for [57].
Photonics 13 00054 g002
Table 1. Operational principles and characteristics of three primary fNIRS methodologies.
Table 1. Operational principles and characteristics of three primary fNIRS methodologies.
FeatureContinuous-WaveFrequency-DomainTime-Domain
Basic PrincipleUses a continuous, constant-intensity light sourceUses a light source at radio frequencies Uses a picosecond pulsed light source
Measured ParametersIntensity AttenuationAmplitude and phase shift in the modulated signal.Distribution of times of flight of photons
Advantage1. Simple system structure;
2. High sampling rate and real-time performance;
3. High SNR for measuring relative changes		1. Capability of measuring absolute optical parameters;
2. Better depth resolution than CW technology;
3. Strong anti-interference ability;		1. Directly measure absolute optical parameters and provides depth resolution;
2. Optimal depth resolution;
3. Wide application range;
Disadvantage1. Inability to measure absorption from scattering;
2. Limited quantitative accuracy;
3. Poor depth resolution;		1. High system complexity;
2. Limited modulation frequency;
3. Complex data processing; 		1. Extremely high system complexity (ultra-short pulse lasers, TCSPC);
2. Slow sampling rate;
3. Stringent environmental requirements;
EvaluationSimple and low-cost but limited informationModerately complex and costlyGold standard but complex and expensive.
Table 2. Comparison and Concordance of fNIRS with Other Brain Imaging Techniques.
Table 2. Comparison and Concordance of fNIRS with Other Brain Imaging Techniques.
Comparison DimensionfNIRS PerformanceConcordance/Difference with Other TechniquesRelevant Studies (Author/Year/Reference)
Spatial ResolutionLimited (~1–2 cm), restricted to the cortexLower than fMRI, but can be improved with high-density arrays or MRI co-registrationCui et al. (2011) [64]; Eggebrecht et al. (2014) [67]
Temporal ResolutionRelatively high (~0.1–1 Hz), suitable for continuous monitoringSuperior to fMRI, but lower than EEGYücel et al. (2021) [73]
Ecological ValidityHigh; portable, silent, suitable for naturalistic settingsSuperior to both fMRI and EEG (in terms of motion tolerance)Piper et al. (2014) [1]; Ayaz et al. (2013) [3]
Clinical ApplicabilitySuitable for infants, patient populations, and multimodal integrationComplements fMRI/EEG in areas such as epilepsy and aphasiaVannasing et al. (2016) [59]; Meier et al. (2023) [57]
Signal ContaminationSusceptible to scalp blood flow, hair colorRequires extra preprocessing; less stable than fMRIKwasa et al. (2023) [68]; Di Lorenzo et al. (2019) [75]
StandardizationLow; diverse analysis pipelines existLack of unified standards hinders cross-study comparabilityYücel et al. (2021) [73]
Table 3. Language-Related Brain Regions and Their Functions Revealed by fNIRS.
Table 3. Language-Related Brain Regions and Their Functions Revealed by fNIRS.
Brain RegionFunctional DescriptionRepresentative Related Studies
(Author/Year/Reference)
Inferior Frontal Gyrus Language production, syntactic processing, speech perceptionAltvater-Mackensen (2018) [33];
Wagley et al. (2024) [45];
Jackson et al. (2021) [53]
Middle Temporal Gyrus/Superior Temporal Gyrus Semantic processing, auditory processing, speech perceptionDefenderfer et al. (2021) [26];
Wagley et al. (2024) [45];
Lawrence et al. (2021) [35]
Prefrontal Cortex/Dorsolateral Prefrontal Cortex Lexical acquisition; cross-linguistic processingZhou et al. (2024) [39];
Li, Y et al. (2020) [43];
Kou et al. (2024) [46]
Motor CortexSpeech imagery, articulation planningSi et al. (2021) [24];
Jackson et al. (2019) [52]
Supplementary Motor Area Language production, naming, repetitionLi et al. (2023) [56]
Angular GyrusCross-linguistic processingJiang et al. (2023) [49]
Parietal CortexVisuospatial attention, language-attention interactionNiu et al. (2024) [38]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Cui, P.; Cui, Y.; Zhang, X.; Zhang, X. Functional Near-Infrared Spectroscopy in Linguistic Research: Recent Advances and Future Perspectives. Photonics 2026, 13, 54. https://doi.org/10.3390/photonics13010054

AMA Style

Cui P, Cui Y, Zhang X, Zhang X. Functional Near-Infrared Spectroscopy in Linguistic Research: Recent Advances and Future Perspectives. Photonics. 2026; 13(1):54. https://doi.org/10.3390/photonics13010054

Chicago/Turabian Style

Cui, Pengke, Yezhi Cui, Xin Zhang, and Xiu Zhang. 2026. "Functional Near-Infrared Spectroscopy in Linguistic Research: Recent Advances and Future Perspectives" Photonics 13, no. 1: 54. https://doi.org/10.3390/photonics13010054

APA Style

Cui, P., Cui, Y., Zhang, X., & Zhang, X. (2026). Functional Near-Infrared Spectroscopy in Linguistic Research: Recent Advances and Future Perspectives. Photonics, 13(1), 54. https://doi.org/10.3390/photonics13010054

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop