KEYWORDS: 3D modeling, Head, Automatic alignment, Photogrammetry, 3D scanning, Point clouds, Cameras, 3D acquisition, 3D image processing, Solid modeling
SignificanceTo effectively apply functional near-infrared spectroscopy (fNIRS)/diffuse optical tomography (DOT) devices, a three-dimensional (3D) model of the position of each optode on a subject’s scalp and the positions of that subject’s cranial landmarks are critical. Obtaining this information accurately in infants, who rarely stop moving, is an ongoing challenge.AimWe propose a smartphone-based registration system that can potentially achieve a full-head 3D scan of a 6-month-old infant instantly.ApproachThe proposed system is remotely controlled by a custom-designed Bluetooth controller. The scanned images can either be manually or automatically aligned to generate a 3D head surface model.ResultsA full-head 3D scan of a 6-month-old infant can be achieved within 2 s via this system. In testing on a realistic but static infant head model, the average Euclidean error of optode position using this device was 1.8 mm.ConclusionsThis low-cost 3D registration system therefore has the potential to permit accurate and near-instant fNIRS/DOT spatial registration.
Recent progress in optoelectronics has made wearable and high-density functional near-infrared spectroscopy (fNIRS) and diffuse optical tomography (DOT) technologies possible for the first time. These technologies have the potential to open new fields of real-world neuroscience by enabling functional neuroimaging of the human cortex at a resolution comparable to fMRI in almost any environment and population. In this perspective article, we provide a brief overview of the history and the current status of wearable high-density fNIRS and DOT approaches, discuss the greatest ongoing challenges, and provide our thoughts on the future of this remarkable technology.
This report is the second part of a comprehensive two-part series aimed at reviewing an extensive and diverse toolkit of novel methods to explore brain health and function. While the first report focused on neurophotonic tools mostly applicable to animal studies, here, we highlight optical spectroscopy and imaging methods relevant to noninvasive human brain studies. We outline current state-of-the-art technologies and software advances, explore the most recent impact of these technologies on neuroscience and clinical applications, identify the areas where innovation is needed, and provide an outlook for the future directions.
KEYWORDS: Head, Data modeling, Image registration, Absorption, Diffuse optical tomography, Databases, Magnetic resonance imaging, Image restoration, Motion models, Medical research
Diffuse optical tomography relies on anatomical models to simulate light transport. We investigate which cotside measures are best to choose an individual-level head model when subject-specific data is unavailable for neonatal infants.
We have developed a series of wearable high-density diffuse optical tomography (HD-DOT) technologies specifically for neonatal applications. These systems provide an ultra-lightweight form factor, a low profile and high mechanical flexibility. This new technology is validated using a novel, anatomically accurate dynamic phantom.
We investigated the performance of a novel HD-DOT system by replicating a series of classic visual stimulation paradigms. Haemodynamic response functions and cortical activation maps replicated the results obtained with larger fibre-based systems.
We applied a wearable 24-module high-density diffuse optical tomography (HD-DOT) system in a resting state (RS) paradigm repeatedly in one subject. Seed-based correlation maps show large field-of-view RS functional connectivity.
We are translating wearable HD-DOT to the neonatal clinic to investigate healthy and brain-injured infants and establish a model of the developmental trajectory of the infant sensorimotor system.
Significance: Early monolingual versus bilingual experience induces adaptations in the development of linguistic and cognitive processes, and it modulates functional activation patterns during the first months of life. Resting-state functional connectivity (RSFC) is a convenient approach to study the functional organization of the infant brain. RSFC can be measured in infants during natural sleep, and it allows to simultaneously investigate various functional systems. Adaptations have been observed in RSFC due to a lifelong bilingual experience. Investigating whether bilingualism-induced adaptations in RSFC begin to emerge early in development has important implications for our understanding of how the infant brain’s organization can be shaped by early environmental factors.
Aims: We attempt to describe RSFC using functional near-infrared spectroscopy (fNIRS) and to examine whether it adapts to early monolingual versus bilingual environments. We also present an fNIRS data preprocessing and analysis pipeline that can be used to reliably characterize RSFC in development and to reduce false positives and flawed results interpretations.
Methods: We measured spontaneous hemodynamic brain activity in a large cohort (N = 99) of 4-month-old monolingual and bilingual infants using fNIRS. We implemented group-level approaches based on independent component analysis to examine RSFC, while providing proper control for physiological confounds and multiple comparisons.
Results: At the group level, we describe the functional organization of the 4-month-old infant brain in large-scale cortical networks. Unbiased group-level comparisons revealed no differences in RSFC between monolingual and bilingual infants at this age.
Conclusions: High-quality fNIRS data provide a means to reliably describe RSFC patterns in the infant brain. The proposed group-level RSFC analyses allow to assess differences in RSFC across experimental conditions. An effect of early bilingual experience in RSFC was not observed, suggesting that adaptations might only emerge during explicit linguistic tasks, or at a later point in development.
KEYWORDS: Diffuse optical tomography, Visualization, Neurophotonics, Visual cortex, Head, Brain, Hemodynamics, 3D modeling, Signal to noise ratio, Imaging systems
Significance: High-density diffuse optical tomography (HD-DOT) has been shown to approach the resolution and localization accuracy of blood oxygen level dependent-functional magnetic resonance imaging in the adult brain by exploiting densely spaced, overlapping samples of the probed tissue volume, but the technique has to date required large and cumbersome optical fiber arrays.
Aim: To evaluate a wearable HD-DOT system that provides a comparable sampling density to large, fiber-based HD-DOT systems, but with vastly improved ergonomics.
Approach: We investigated the performance of this system by replicating a series of classic visual stimulation paradigms, carried out in one highly sampled participant during 15 sessions to assess imaging performance and repeatability.
Results: Hemodynamic response functions and cortical activation maps replicate the results obtained with larger fiber-based systems. Our results demonstrate focal activations in both oxyhemoglobin and deoxyhemoglobin with a high degree of repeatability observed across all sessions. A comparison with a simulated low-density array explicitly demonstrates the improvements in spatial localization, resolution, repeatability, and image contrast that can be obtained with this high-density technology.
Conclusions: The system offers the possibility for minimally constrained, spatially resolved functional imaging of the human brain in almost any environment and holds particular promise in enabling neuroscience applications outside of the laboratory setting. It also opens up new opportunities to investigate populations unsuited to traditional imaging technologies.
Significance: Neonates are a highly vulnerable population. The risk of brain injury is greater during the first days and weeks after birth than at any other time of life. Functional neuroimaging that can be performed longitudinally and at the cot-side has the potential to improve our understanding of the evolution of multiple forms of neurological injury over the perinatal period. However, existing technologies make it very difficult to perform repeated and/or long-duration functional neuroimaging experiments at the cot-side.
Aim: We aimed to create a modular, high-density diffuse optical tomography (HD-DOT) technology specifically for neonatal applications that is ultra-lightweight, low profile and provides high mechanical flexibility. We then sought to validate this technology using an anatomically accurate dynamic phantom.
Approach: An advanced 10-layer rigid-flexible printed circuit board technology was adopted as the basis for the DOT modules, which allows for a compact module design that also provides the flexibility needed to conform to the curved infant scalp. Two module layouts were implemented: dual-hexagon and triple-hexagon. Using in-built board-to-board connectors, the system can be configured to provide a vast range of possible layouts. Using epoxy resin, thermochromic dyes, and MRI-derived 3D-printed moulds, we constructed an electrically switchable, anatomically accurate dynamic phantom. This phantom was used to quantify the imaging performance of our flexible, modular HD-DOT system.
Results: Using one particular module configuration designed to cover the infant sensorimotor system, the device provided 36 source and 48 detector positions, and over 700 viable DOT channels per wavelength, ranging from 10 to ∼45 mm over an area of approximately 60 cm2. The total weight of this system is only 70 g. The signal changes from the dynamic phantom, while slow, closely simulated real hemodynamic response functions. Using difference images obtained from the phantom, the measured 3D localization error provided by the system at the depth of the cortex was in the of range 3 to 6 mm, and the lateral image resolution at the depth of the neonatal cortex is estimated to be as good as 10 to 12 mm.
Conclusions: The HD-DOT system described is ultra-low weight, low profile, can conform to the infant scalp, and provides excellent imaging performance. It is expected that this device will make functional neuroimaging of the neonatal brain at the cot-side significantly more practical and effective.
We introduce a new wearable HD-DOT system that allows neuroimaging in naturalistic environments. Test results with visual paradigms show comparable performance to larger fiber-based systems.
The application of functional near-infrared spectroscopy (fNIRS) in the neurosciences has been expanding over the last 40 years. Today, it is addressing a wide range of applications within different populations and utilizes a great variety of experimental paradigms. With the rapid growth and the diversification of research methods, some inconsistencies are appearing in the way in which methods are presented, which can make the interpretation and replication of studies unnecessarily challenging. The Society for Functional Near-Infrared Spectroscopy has thus been motivated to organize a representative (but not exhaustive) group of leaders in the field to build a consensus on the best practices for describing the methods utilized in fNIRS studies.
Our paper has been designed to provide guidelines to help enhance the reliability, repeatability, and traceability of reported fNIRS studies and encourage best practices throughout the community. A checklist is provided to guide authors in the preparation of their manuscripts and to assist reviewers when evaluating fNIRS papers.
Critical glycemic events, such as hypo- or hyperglycemia, are extremely common during the first week post-partum in very preterm neonates. Both hypo- and hyperglycemic changes have been associated with poor neurological outcome. Continuous glucose monitoring (CGM) is a promising tool to reduce glycemic variability in the preterm population and whole-head Diffuse Optical Tomography (DOT) is a promising tool for continuous monitoring of brain hemodynamics in newborns. In this study, we performed a combined CGM-DOT acquisition in a very preterm newborn (28 weeks gestational age). The newborn was monitored for 7 days continuously. Twelve events were detected during this period: 8 mild hypoglycemic events, one severe hypoglycemic event, two mild hyperglycemic events and one event with a mild hypo- followed by a mild-hyperglycemia. DOT data were available for all the events but two. DOT data were reconstructed with a neonatal head model for the severe hypoglycemic event before the start of the hypoglycemic event and during the maximum peak of hypoglycemia. These preliminary results showed a regional specificity of the hemodynamic changes during hypoglycemia, with a predominant recruitment of the motor and parietal areas. This study highlights the importance of using whole-head DOT in this research field and the feasibility to perform combined CGMDOT monitoring in very preterm neonates. Future clinical trials are required to investigate this clinical problem more thoroughly and shed light on the impact of tight glycemic control on the newborn brain.
The next generation of diffuse optical imaging systems will consist of wearable and fiber-less devices, to exploit the advantages of diffuse optical imaging over other functional neuroimaging techniques and meet the needs of users to acquire data in real-world settings. Recently, research at UCL gave rise to a novel, modular high-density diffuse optical tomography (DOT) system that was validated by reconstructing activation images over the motor cortex of a thumb-tofinger extension task. The real question, however, is whether these fiber-less systems can be employed whilst the subject performs real-world activities, that is, whether they can provide reliable signals during participant motion. Integrating motion sensors into modular wearable electronics is straightforward. In this study we acquired DOT and motion sensor data whilst participants performed different activities involving motion. In one acquisition, only accelerometer data were acquired while in the second acquisition, all 9-axis of data (accelerometer, gyroscope and magnetometer data) were acquired. Results demonstrated that acceleration data from motion sensors is not enough to detect motion artifacts whilst performing active movement (e.g., walking), since the global motion obscures any subtle motion artifact. Conversely, by combining accelerometer and gyroscope data it seems possible to detect motion artifacts even during walking, that is when a global motion is present. However, not all types of motion artifacts (e.g., eyebrow raising) could be detected even with this full data configuration. Further studies are required to shed light on this important research question.
We discuss advances in and applications of fibre-less, wearable, high-density diffuse optical tomography technologies, including a new device specifically for the newborn infant that employs flex-rigid PCB technology and provides channel density approaching 10 channels/cm2.
The position of each source and detector “optode” on the scalp, and their relative separations, determines the sensitivity of each functional near-infrared spectroscopy (fNIRS) channel to the underlying cortex. As a result, selecting appropriate scalp locations for the available sources and detectors is critical to every fNIRS experiment. At present, it is standard practice for the user to undertake this task manually; to select what they believe are the best locations on the scalp to place their optodes so as to sample a given cortical region-of-interest (ROI). This process is difficult, time-consuming, and highly subjective. Here, we propose a tool, Array Designer, that is able to automatically design optimized fNIRS arrays given a user-defined ROI and certain features of the available fNIRS device. Critically, the Array Designer methodology is generalizable and will be applicable to almost any subject population or fNIRS device. We describe and validate the algorithmic methodology that underpins Array Designer by running multiple simulations of array design problems in a realistic anatomical model. We believe that Array Designer has the potential to end the need for manual array design, and in doing so save researchers time, improve fNIRS data quality, and promote standardization across the field.
The development of a whole-scalp, high sampling-density diffuse optical tomography (DOT) system is a critical next step in the evolution of the field of diffuse optics. To achieve this with optical fiber bundles is extremely challenging, simply because of the sheer number of bundles required, and the associated challenges of weight and ergonomics. Dispensing with optical fiber bundles and moving to head-mounted optoelectronics can potentially facilitate the advent of a new generation of wearable, whole-scalp technologies that will open up a range of new experimental and clinical applications for diffuse optical measurements. Here, we present a concise review of the significant progress that has been made toward achieving a wearable, fiberless, high-density, whole-scalp DOT system. We identify the key limitations of current technologies and discuss the possible opportunities for future development.
In diffuse optical tomography (DOT), overlapping and multidistance measurements are required to reconstruct depth-resolved images of oxy- (HbO2) and deoxy- (HHb) hemoglobin concentration changes occurring in the brain. These can be considered an indirect measure of brain activity, under the assumption of intact neurovascular coupling. Broadband systems also allow changes in the redox state of cytochrome c oxidase (oxCCO) to be measured, which can be an important biomarker when neurovascular coupling is impaired. We used DOT to reconstruct images of Δ[HbO2], Δ[HHb], and Δ[oxCCO] from data acquired with a broadband system. Four healthy volunteers were measured while performing a visual stimulation task (4-Hz inverting checkerboard). The broadband system was configured to allow multidistance and overlapping measurements of the participants’ visual cortex with 32 channels. A multispectral approach was employed to reconstruct changes in concentration of the three chromophores during the visual stimulation. A clear and focused activation was reconstructed in the left occipital cortex of all participants. The difference between the residuals of the three-chromophore model and of the two-chromophore model (recovering only Δ[HbO2] and Δ[HHb]) exhibits a spectrum similar to that of oxCCO. These results form a basis for further studies aimed to further optimize image reconstruction of Δ[oxCCO].
Burst suppression (BS) is an electroencephalographic state associated with a profound inactivation of the brain. BS and pathological discontinuous electroencephalography (EEG) are often observed in term-age infants with neurological injury and can be indicative of a poor outcome and lifelong disability. Little is known about the neurophysiological mechanisms of BS or how the condition relates to the functional state of the neonatal brain. We used simultaneous EEG and diffuse optical tomography (DOT) to investigate whether bursts of EEG activity in infants with hypoxic ischemic encephalopathy are associated with an observable cerebral hemodynamic response. We were able to identify significant changes in concentration of both oxy and deoxyhemoglobin that are temporally correlated with EEG bursts and present a relatively consistent morphology across six infants. Furthermore, DOT reveals patient-specific spatial distributions of this hemodynamic response that may be indicative of a complex pattern of cortical activation underlying discontinuous EEG activity that is not readily apparent in scalp EEG.
We present a method for acquiring whole-head images of changes in blood volume and oxygenation from the infant brain at cot-side using time-resolved diffuse optical tomography (TR-DOT). At UCL, we have built a portable TR-DOT device, known as MONSTIR II, which is capable of obtaining a whole-head (1024 channels) image sequence in 75 seconds. Datatypes extracted from the temporal point spread functions acquired by the system allow us to determine changes in absorption and reduced scattering coefficients within the interrogated tissue. This information can then be used to define clinically relevant measures, such as oxygen saturation, as well as to reconstruct images of relative changes in tissue chromophore concentration, notably those of oxy- and deoxyhaemoglobin. Additionally, the effective temporal resolution of our system is improved with spatio-temporal regularisation implemented through a Kalman filtering approach, allowing us to image transient haemodynamic changes. By using this filtering technique with intensity and mean time-of-flight datatypes, we have reconstructed images of changes in absorption and reduced scattering coefficients in a dynamic 2D phantom. These results demonstrate that MONSTIR II is capable of resolving slow changes in tissue optical properties within volumes that are comparable to the preterm head. Following this verification study, we are progressing to imaging a 3D dynamic phantom as well as the neonatal brain at cot-side. Our current study involves scanning healthy babies to demonstrate the quality of recordings we are able to achieve in this challenging patient population, with the eventual goal of imaging functional activation and seizures.
In recent years, it has been demonstrated that using functional near-infrared spectroscopy (fNIRS) channels with short separations to explicitly sample extra-cerebral tissues can provide a significant improvement in the accuracy and reliability of fNIRS measurements. The aim of these short-separation channels is to measure the same superficial hemodynamics observed by standard fNIRS channels while also being insensitive to the brain. We use Monte Carlo simulations of photon transport in anatomically informed multilayer models to determine the optimum source–detector distance for short-separation channels in adult and newborn populations. We present a look-up plot that provides (for an acceptable value of short-separation channel brain sensitivity relative to standard channel brain sensitivity) the optimum short-separation distance. Though values vary across the scalp, when the acceptable ratio of the short-separation channel brain sensitivity to standard channel brain sensitivity is set at 5%, the optimum short-separation distance is 8.4 mm in the typical adult and 2.15 mm in the term-age infant.
Functional near-infrared spectroscopy (fNIRS) is an optical imaging method that is used to noninvasively measure cerebral hemoglobin concentration changes induced by brain activation. Using structural guidance in fNIRS research enhances interpretation of results and facilitates making comparisons between studies. AtlasViewer is an open-source software package we have developed that incorporates multiple spatial registration tools to enable structural guidance in the interpretation of fNIRS studies. We introduce the reader to the layout of the AtlasViewer graphical user interface, the folder structure, and user files required in the creation of fNIRS probes containing sources and detectors registered to desired locations on the head, evaluating probe fabrication error and intersubject probe placement variability, and different procedures for estimating measurement sensitivity to different brain regions as well as image reconstruction performance. Further, we detail how AtlasViewer provides a generic head atlas for guiding interpretation of fNIRS results, but also permits users to provide subject-specific head anatomies to interpret their results. We anticipate that AtlasViewer will be a valuable tool in improving the anatomical interpretation of fNIRS studies.
Optical imaging techniques provide a means of monitoring haemodynamics and tissue oxygenation by virtue of the differing absorption spectra of relevant endogenous chromophores. Whilst time-domain diffuse optical tomography offers sufficient sensitivity to produce full three dimensional images of such properties through the entire infant brain, standard approaches to the imaging protocol and reconstruction methods limit the temporal resolution which can be achieved without an unacceptable degradation in the image quality. In this work we employ spatio-temporal regularisation by means of a variational form Kalman filter to achieve significantly improved temporal resolution whilst maintaining image quality. We demonstrate this approach in a dynamic phantom study where we successfully track moving absorbing and scattering targets using the MONSTIR II instrument developed at University College London.
The production of accurate and independent images of the changes in concentration of oxyhemoglobin and deoxyhemoglobin by diffuse optical imaging is heavily dependent on which wavelengths of near-infrared light are chosen to interrogate the target tissue. Although wavelengths can be selected by theoretical methods, in practice the accuracy of reconstructed images will be affected by wavelength-specific and system-specific factors such as laser source power and detector sensitivity. We describe the application of a data-driven approach to optimum wavelength selection for the second generation of University College London’s multichannel, time-domain optical tomography system (MONSTIR II). By performing a functional activation experiment using 12 different wavelengths between 690 and 870 nm, we were able to identify the combinations of 2, 3, and 4 wavelengths which most accurately reproduced the results obtained using all 12 wavelengths via an imaging approach. Our results show that the set of 2, 3, and 4 wavelengths which produce the most accurate images of functional activation are [770, 810], [770, 790, 850], and [730, 770, 810, 850] respectively, but also that the system is relatively robust to wavelength selection within certain limits. Although these results are specific to MONSTIR II, the approach we developed can be applied to other multispectral near-infrared spectroscopy and optical imaging systems.
Performance assessment of instruments devised for clinical applications is of key importance for validation and quality assurance. Two new protocols were developed and applied to facilitate the design and optimization of instruments for time-domain optical brain imaging within the European project nEUROPt. Here, we present the “Basic Instrumental Performance” protocol for direct measurement of relevant characteristics. Two tests are discussed in detail. First, the responsivity of the detection system is a measure of the overall efficiency to detect light emerging from tissue. For the related test, dedicated solid slab phantoms were developed and quantitatively spectrally characterized to provide sources of known radiance with nearly Lambertian angular characteristics. The responsivity of four time-domain optical brain imagers was found to be of the order of 0.1 m2 sr. The relevance of the responsivity measure is demonstrated by simulations of diffuse reflectance as a function of source-detector separation and optical properties. Second, the temporal instrument response function (IRF) is a critically important factor in determining the performance of time-domain systems. Measurements of the IRF for various instruments were combined with simulations to illustrate the impact of the width and shape of the IRF on contrast for a deep absorption change mimicking brain activation.
An optical imaging system has been developed which uses measurements of diffusely reflected near-infrared light to
produce maps of changes in blood flow and oxygenation occurring within the cerebral cortex. Optical sources and
detectors are coupled to the head via an array of optical fibers, on a probe held in contact with the scalp, and data is
collected at a rate of 10 Hz. A clinical electroencephalography (EEG) system has been integrated with the optical system
to enable simultaneous observation of electrical and hemodynamic activity in the cortex of neurologically compromised
newborn infants diagnosed with seizures. Studies have made a potentially critically important discovery of previously
unknown transient hemodynamic events in infants treated with anticonvulsant medication. We observed repeated
episodes of small increases in cortical oxyhemoglobin concentration followed by a profound decrease in 3 of 4 infants
studied, each with cerebral injury who presented with neonatal seizures. This was not accompanied by clinical or EEG
seizure activity and was not present in nineteen matched controls. The underlying cause of these changes is currently
unknown. We tentatively suggest that our results may be associated with a phenomenon known as cortical spreading
depolarization, not previously observed in the infant brain.
KEYWORDS: Near infrared spectroscopy, Veins, Monte Carlo methods, Brain, In vivo imaging, Tissue optics, Functional magnetic resonance imaging, Tissues, Oxygen, Data modeling
Near-Infrared Spectroscopy (NIRS) measures the functional hemodynamic response occuring at the surface of
the cortex. Large pial veins are located above the surface of the cerebral cortex. Following activation, these
veins exhibit oxygenation changes but their volume likely stays constant. The back-reflection geometry of the
NIRS measurement renders the signal very sensitive to these superficial pial veins. As such, the measured NIRS
signal contains contributions from both the cortical region as well as the pial vasculature. In this work, the
cortical contribution to the NIRS signal was investigated using (1) Monte Carlo simulations over a realistic
geometry constructed from anatomical and vascular MRI and (2) multimodal NIRS-BOLD recordings during
motor stimulation. A good agreement was found between the simulations and the modeling analysis of in vivo
measurements. Our results suggest that the cortical contribution to the deoxyhemoglobin signal change (ΔHbR)
is equal to 16-22% of the cortical contribution to the total hemoglobin signal change (ΔHbT). Similarly, the
cortical contribution of the oxyhemoglobin signal change (ΔHbO) is equal to 73-79% of the cortical contribution
to the ΔHbT signal. These results suggest that ΔHbT is far less sensitive to pial vein contamination and
therefore, it is likely that the ΔHbT signal provides better spatial specificity and should be used instead of
ΔHbO or ΔHbR to map cerebral activity with NIRS. While different stimuli will result in different pial vein
contributions, our finger tapping results do reveal the importance of considering the pial contribution.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.