Search

Test-retest reliability of neuroimaging measurements is an important concern in the investigation of cognitive functions in the human brain. To date, intraclass correlation coefficients (ICCs), originally used in interrater reliability studies in behavioral sciences, have become commonly used metrics in reliability studies on neuroimaging and functional near-infrared spectroscopy (fNIRS). However, as there are six popular forms of ICC, the adequateness of the comprehensive understanding of ICCs will affect how one may appropriately select, use, and interpret ICCs toward a reliability study. We first offer a brief review and tutorial on the statistical rationale of ICCs, including their underlying analysis of variance models and technical definitions, in the context of assessment on intertest reliability. Second, we provide general guidelines on the selection and interpretation of ICCs. Third, we illustrate the proposed approach by using an actual research study to assess interest reliability of fNIRS-based, volumetric diffuse optical tomography of brain activities stimulated by a risk decision-making protocol. Last, special issues that may arise in reliability assessment using ICCs are discussed and solutions are suggested.

Sensorimotor cortex plasticity induced by constraint-induced movement therapy (CIMT) in six children (10.2±2.1 years old) with hemiplegic cerebral palsy was assessed by functional near-infrared spectroscopy (fNIRS). The activation laterality index and time-to-peak/duration during a finger-tapping task and the resting-state functional connectivity were quantified before, immediately after, and 6 months after CIMT. These fNIRS-based metrics were used to help explain changes in clinical scores of manual performance obtained concurrently with imaging time points. Five age-matched healthy children (9.8±1.3 years old) were also imaged to provide comparative activation metrics for normal controls. Interestingly, the activation time-to-peak/duration for all sensorimotor centers displayed significant normalization immediately after CIMT that persisted 6 months later. In contrast to this improved localized activation response, the laterality index and resting-state connectivity metrics that depended on communication between sensorimotor centers improved immediately after CIMT, but relapsed 6 months later. In addition, for the subjects measured in this work, there was either a trade-off between improving unimanual versus bimanual performance when sensorimotor activation patterns normalized after CIMT, or an improvement occurred in both unimanual and bimanual performance but at the cost of very abnormal plastic changes in sensorimotor activity.

Additions of a solute/carbohydrate in tissue affect the size of tissue cells and the refractive indexes of the extra- and intracellular fluids, and thus the overall tissue scattering properties. We use both the Rayleigh- Gans and Mie theory approximation in calculating effects of the osmolarity and refractive indexes on the reduced scattering coefficient of tissue, and employ photon diffusion theory to associate the reduced scattering coefficient to the mean optical path length. The calculations show that changes of scattering in tissue depend not only on the change in extracellular refractive index but also on the change in osmolarity, and thus on the change in cell size and volume fraction. Experimentally, we have utilized time-domain and frequencydomain NIR techniques to measure the changes of optical properties caused by an addition of a solute in tissue models and in perfused rat livers. The temperature-dependent path length measurement of the perfused liver confirms the dependence of tissue scattering on the tissue cell size. The results obtained from the liver with three kinds of carbohydrate perfusion display different scattering aspects and can be well explained by changes in cell size and in extracellular as well as intracellular refractive indexes. The consistency between the theoretical and experimental results confirms the dependence of optical properties in (liver) tissue on both tissue osmolarity and relative refractive indexes between the extracellular and intracellular compartments. This study suggests that the NIR technique is a novel and useful tool for noninvasive, physiological monitoring.

Functional near-infrared spectroscopy (fNIRS) is recently utilized as a new approach to assess resting-state functional connectivity (RSFC) in the human brain. For any new technique or new methodology, it is necessary to be able to replicate similar experiments using different instruments in order to establish its liability and reproducibility. We apply two different diffuse optical tomographic (DOT) systems (i.e., DYNOT and CW5), with various probe arrangements to evaluate RSFC in the sensorimotor cortex by utilizing a previously published experimental protocol and seed-based correlation analysis. Our results exhibit similar spatial patterns and strengths in RSFC between the bilateral motor cortexes. The consistent observations are obtained from both DYNOT and CW5 systems, and are also in good agreement with the previous fNIRS study. Overall, we demonstrate that the fNIRS-based RSFC is reproducible by various DOT imaging systems among different research groups, enhancing the confidence of neuroscience researchers and clinicians to utilize fNIRS for future applications.

We demonstrate the utility of functional near-infrared spec- troscopy fNIRS as a tool for physicians to study cortical plasticity in children with cerebral palsy CP. Motor cortex activation patterns were studied in five healthy children and five children with CP 8.4±2.3 years old in both groups performing a finger-tapping pro- tocol. Spatial distance from center and area difference and temporal duration and time-to-peak image metrics are proposed as potential biomarkers for differentiating abnormal cortical activation in children with CP from healthy pediatric controls. In addition, a similarity image-analysis concept is presented that unveils areas that have simi- lar activation patterns as that of the maximum activation area, but are not discernible by visual inspection of standard activation images. Metrics derived from the images presenting areas of similarity are shown to be sensitive identifiers of abnormal activation patterns in children with CP. Importantly, the proposed similarity concept and related metrics may be applicable to other studies for the identifica- tion of cortical activation patterns by fNIRS. © 2010 Society of Photo-Optical

Surgeons often cannot see major vessels embedded in adipose tissue and inadvertently injure them. One such example occurs during surgical removal of the gallbladder, where injury of the nearby common bile duct leads to life-threatening complications. Near-infrared imaging of the intraoperative field may help surgeons localize such critical tissue-embedded vessels. We have investigated how continuous-wave (CW) imaging performs relative to time-gated wide-field imaging, presently a rather costly technology, under broad Gaussian beam-illumination conditions. We have studied the simplified case of an isolated cylinder having bile-duct optical properties, embedded at different depths within a 2-cm slab of adipose tissue. Monte Carlo simulations were preformed for both reflectance and transillumination geometries. The relative performance of CW versus time-gated imaging was compared in terms of spatial resolution and contrast-to-background ratio in the resulting simulated images. It was found that time-gated imaging offers superior spatial resolution and vessel-detection sensitivity in most cases, though CW transillumination measurements may also offer satisfactory performance for this tissue geometry at lower cost. Experiments were performed in reflectance geometry to validate simulation results, and potential challenges in the translation of this technology to the clinic are discussed.

Our goal is to quantify scattering properties of near-IR light in the rat spinal cord region and to differentiate healthy and demyeli- nated peripheral nerves intraoperatively based on differential light scattering. For the rat spinal cord, optical reflectance is measured from the spinal cord surface at spatial intervals of 1m musing a needle probe. Data are acquired from left and right lumbar regions of the animals as well as on the central blood vessels. The reduced scatter- ing coefficient s is found to be higher 34.2±2.1 cm 1 in the lum- bar regions of the spinal cord than on the central blood vessel 19.9±1.0 cm 1 . This methodology is extended to detect differences in the rat sciatic nerves following left L4 spinal nerve ligation. The reflectance is taken at the same five regions at postoperative days 1, 4, 7, and 14. Significant differences are seen in both the spectral slope and s values on postoperative days 4, 7, and 14, indicating that either of the two quantities could be used as a marker for demyelina- tion. We prove the usefulness of the technique, which may have a possible clinical application for minimally invasive, intraoperative di- agnosis and monitoring of demyelination diseases, such as multiple sclerosis in the central nervous system or degeneration of the periph- eral nervous system. © 2005 Society of Photo-Optical Instrumentation Engineers. DOI: 10.1117/1.2098487

Repetitive transcranial magnetic stimulation (rTMS) is a technology that stimulates neurons with rapidly changing magnetic pulses with demonstrated therapeutic applications for various neuropsychiatric disorders. Functional near-infrared spectroscopy (fNIRS) is a suitable tool to assess rTMS-evoked brain responses without interference from the magnetic or electric fields generated by the TMS coil. We have previously reported a channel-wise study of combined rTMS/fNIRS on the motor and prefrontal cortices, showing a robust decrease of oxygenated hemoglobin concentration (Δ[HbO2]) at the sites of 1-Hz rTMS and the contralateral brain regions. However, the reliability of this putative clinical tool is unknown. In this study, we develop a rapid optical topography approach to spatially characterize the rTMS-evoked hemodynamic responses on a standard brain atlas. A hemispherical approximation of the brain is employed to convert the three-dimensional topography on the complex brain surface to a two-dimensional topography in the spherical coordinate system. The test-retest reliability of the combined rTMS/fNIRS is assessed using repeated measurements performed two to three days apart. The results demonstrate that the Δ[HbO2] amplitudes have moderate-to-high reliability at the group level; and the spatial patterns of the topographic images have high reproducibility in size and a moderate degree of overlap at the individual level.

We hypothesize that nonlinearity between short-term anagram tasks and corresponding hemodynamic responses can be observed by functional near-infrared spectroscopy (fNIRS) in the prefrontal cortex (PFC). The PFC of six human subjects in response to anagram tasks is investigated using multichannel fNIRS. Concentration changes of oxyhemoglobin and deoxyhemoglobin in the PFC are measured with variable anagram durations and at two difficulty levels (four- and six-letter anagrams). The durations to perform the selected anagram tasks range from several seconds to more than one minute. The dorsolateral PFC areas exhibit consistent and strong hemodynamic deactivation during and shortly after task execution. The superposition principle of a linear system is employed to investigate nonlinear hemodynamic features among three task duration subgroups: D1 = 2.0 sec, D2 = 4.0 sec, and D3 = 8.0 sec. Such analysis shows clear nonlinearity in hemodynamic responses on the PFC with task durations shorter than 4 sec. Our observation of significant deactivation in early hemodynamic responses in the PFC is consistent with multiple fNIRS studies and several reports given in the field of functional magnetic resonance imaging. A better understanding of nonlinearity in fNIRS signals will have potential for us to investigate brain adaptation and to extrapolate neuronal activities from hemodynamic signals.

This study investigates the correlation of tumor blood oxygenation and tumor pO(2) with respect to carbogen inhalation. After having refined and validated the algorithms for calculating hemoglobin concentrations, we used near-infrared spectroscopy (NIRS) to measure changes of oxygenated hemoglobin concentration (delta[HbO(2)]) and used an oxygen needle electrode and (19)F MRI for pO(2) measurements in tumors. The measurements were taken from Dunning prostate R3327 tumors implanted in rats, while the anesthetized rats breathed air or carbogen. The NIRS results from tumor measurements showed significant changes in tumor vascular oxygenation in response to carbogen inhalation, while the pO(2) electrode results showed an apparent heterogeneity for tumor pO(2) response to carbogen inhalation, which was also confirmed by (19)F MR pO(2) mapping. Furthermore, we developed algorithms to estimate hemoglobin oxygen saturation, sO(2), during gas intervention based on the measured values of delta[HbO(2)] and pO(2). The algorithms have been validated through a tissue-simulating phantom and used to estimate the values of sO(2) in the animal tumor measurement based on the NIRS and global mean pO(2) values. This study demonstrates that the NIRS technology can provide an efficient, real-time, noninvasive approach to monitoring tumor physiology and is complementary to other techniques, while it also demonstrates the need for an NIR imaging technique to study spatial heterogeneity of tumor vasculature under therapeutic interventions.

Parkinson's disease (PD) is a chronic, progressive disease involving the globus pallidus (GP), which is a gray matter mass, surrounded by white matter deep within the brain. During a neurosurgery procedure, a thin probe is inserted into the GP to create a lesion that often relieves the cardinal symptoms of PD. The goal of this study is to develop an optical method to accurately locate the GP border. In theory, Monte Carlo simulations were performed to predict the optical reflectance from brain tissue. In experiment, a portable, real-time display spectrometer with a fiber optic reflectance probe was developed and used during human surgery. Optical reflectance values were recorded at 1 mm intervals to obtain a spatial profile of the tissue as the probe passed through regions of gray and white matter. The simulation and in vivo studies of the reflectance from the brain are in good agreement with one another. The clinical data show that the reflectance from gray matter is approximately 50% or less than that from white matter between 650 and 800 nm. A slope algorithm is developed to distinguish gray and white matter in vivo. This study provides previously unknown optical reflectance of the human brain. © 1998 Society of Photo-Optical Instrumentation Engineers.