Open Access
Add to BibSonomy
Abstract
The present work shows the capability of near infrared (NIR) light to reach the cerebral cortex through the frontal sinus using continuous-wave techniques (CW-DOT) in a dual study. On the one hand, changes in time during the tracking of a blood dye in the prefrontal cortex were monitored. On the other hand, hemodynamic changes induced by low frequency of transcranial magnetic stimulation applied on the prefrontal cortex were recorded. The results show how NIR light projected through the frontal sinus reaches the cerebral cortex target, providing enough information to have a reliable measurement of cortical hemodynamic changes using CW-DOT.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Diffuse optical tomography (DOT) is a noninvasive imaging technique which measures changes in the absorption of NIR light to detect functional changes during cerebral activations. DOT allows an estimation of changes in cerebral oxygenated (HbO) and deoxygenated hemoglobin (HbR) concentrations due to local brain activation [1]. In order to do this, DOT uses at least two wavelengths in a range of 650-950 nm which measure the absorbed quantity of NIR light from biological tissue by measuring diffusely scattered light [2]. Diffuse properties of the light are limited, but are sufficient to study most of the cerebral cortex with a penetration of around 3-4 cm in human brain imaging. The NIR light is applied on the subject’s head through LEDs or optical fibers which are combined as source-detectors to detect changes in the optical density produced by hemodynamic changes in the brain [3].The changes in NIR light attenuation between source and detector located on the scalp are transformed into changes in the concentration of HbO and HbR by applying the modified Beer-Lambert law [4], which assumes the scattering constant in the time and that changes in detected light intensity are attributed to changes in HbO and HbR [5]. Moreover, the multidistance approach used by DOT transforms the detected light from different measuring distances on the surface of the head into deep information providing three-dimensional images of cerebral activations versus the topographic approach which provides planar back-projection, allowing an increase in spatial resolution and positional accuracy of optical brain imaging [6].
In order to create functional images, the NIR light needs to pass through superficial layers such as the scalp and skull, both before and after passing through the brain tissues [7]. Due to the heterogeneity in the optical properties of the tissue [8], the light propagation in the head could be affected by anatomical structures. The more sophisticated models of light propagation in the head use anatomical images from magnetic resonance images (MRI) to model the light propagation in each layer of the head. Generally, the models of light propagation in the head are simplified and they usually consist of five layers formed by the scalp, skull, cerebrospinal fluid, white matter and gray matter [9,10].
From an anatomical point of view, the five layers selected could be insufficient for the light propagation modeling in the head, because there are anatomical structures that can disturb the light behavior during the travel through layers of the head such as the subarachnoid space. It has previously been shown, in the subarachnoid space, that low scattering properties do not drastically affect the light propagation in the medium [11]. The frontal sinus (FS) is an anatomical structure which is not usually taken into account during light propagation modeling in the head. NIR light should cross the FS until the cerebral cortex, when performing functional studies which involve the prefrontal cortex. Therefore, knowing whether or not it is possible to reach the frontal cortex is a basic aspect of this research.
The bibliography just shows simulation studies of light propagation across the FS [12] [7,13], but they are insufficient to prove whether NIR light can reach the prefrontal cortex. The morphology of the FS is individually highly variable [14,15]. Uneven distribution of superficial tissue layer thickness in the head affects the mapping of hemodynamic changes in the frontal area, therefore the changes in NIR light absorption are variable depending on the scalp-cortex distance [15,16].
Cerebral blood inflow monitoring may help to know whether NIR light is able to measure hemodynamic changes in the brain through the FS in a real space without using simulation models. The tracking of the cerebral blood inflow using contrast agents or dyes are methods that provide information on the time signal changes. Previous studies have evaluated the separation of intra and extra-cerebral fractions of the signals using indocyanine green (ICG) which absorbs light in the near infrared spectrum. Some depth resolved NIRS techniques such as time-domain (TD) [17], frequency-domain (FD) [18] and continuous wave (CW) [19] have been used to measure changes in the NIR signals during the ICG inflow in the motor cortex [19] and occipital cortex [18]. If the ICG tracking is used to separate extracerebral and intracerebral signals on the prefrontal cortex using CW-DOT as monitoring device, it is possible to study the reliability of NIR light reaching the cerebral cortex target behind the FS. Therefore, one aim of the present work is to study the influence of the FS in hemodynamic measures DOT using ICG as marked cerebral blood, which has not been done until now.
Due to the fact that ICG tracking on the prefrontal cortex using DOT devices to corroborate the signal changes in the time measured by CW-DOT has not been reported previously, another imaging technique is necessary as control. Dynamic magnetic susceptibility imaging relies on the rapid acquisition of as many images as possible during the passage of contrast media through the brain to measure the degree of the time signal changes [20].Contrast media tracking in MR is a similar technique to ICG tracking, allowing the comparison of the changes of both signals in time on the prefrontal cortex behind the FS.
Furthermore, noninvasive methods based on cognitive tasks allow the measurement of hemodynamic changes at the prefrontal cortex level and the reconstruction of functional DOT images using a CW-DOT device. However, if the intention is to show the capability and reliability of NIR light reaching the cerebral cortex through the FS, the results based on cognitive tasks may not be reliable because the cerebral activations on the prefrontal cortex are more differential and subtle than cerebral activations generated by other tasks [21]. In addition, the influence of the FS has not been well established using a tomography approach whose localization of the cerebral activations in depth could be affected.
However, there is a method that could guarantee the cerebral activations in a controller manner, which could be the transcranial magnetic stimulation (TMS). TMS generates activity changes at the cerebral cortex level produced by a magnetic pulse passing a brief electric current through a magnetic coil [22]. A CW-DOT device can monitor hemodynamic changes generated by the TMS, simultaneously. As a result, reconstructed 3D activity images can be recorded by CW-DOT during a repetitive transcranial magnetic stimulation (rTMS) and used to study the reliability of the NIR light reaching the cerebral cortex target behind the FS, which is the goal of the present work.
2. Material and methods
2.1. Subjects
Two healthy volunteers (subject A and subject B) with no history of any neurological disease participated in the dual study. Written informed consent was explained and signed prior to the experiments. The study was approved by the local ethics committee (Universidad de La Laguna) and was conducted in accordance with the Declaration of Helsinki. A dual study was conducted based on: a) cerebral blood inflow monitoring using CW-DOT system and MRI devices on subject A, and b) hemodynamic changes recording generated by rTMS on subject B.
2.2. Procedure of the inflow of a contrast agent in MRI
A magnetic resonance (MR) scanner was used to monitor cerebral blood inflow in the prefrontal cortex of subject A. T2*BOLD images were acquired from an MR device for monitoring the changes in the time magnetic signal intensity produced by a contrast agent based on gadolinium (Gd-DO3A-butrol, Gd) [23]. The first 5 sec of the scan were acquired before contrast agent injection to establish a pre-contrast baseline. Then, a bolus of 5.5 ml of contrast agent was injected into the cubital vein of the right arm followed by a bolus of 5 ml of saline.
2.3. Data acquisition in MRI
Changes in magnetic susceptibly during the passage of Gd through the capillary bed were measured by a 3T Signa Excite HD scanner (General Electric). A sequence of perfusion based on T2*-weighted echo-planar imaging of 256 volumes was acquired; 8 axial slices covered the prefrontal area; field of view 26 mm, slice thickness 4 mm, inter-slice gap 1 mm, 96 x 128 matrix, flip angle 49°, TR of 552 msec, were acquired to measure the changes in the magnetic signal intensity during the passage of the contrast agent. T1-weighted volume was acquired for precise anatomical localization (time repetition (TR) = 6 msec, time echo (TE) = 1 msec, flip angle = 12°, matrix size = 256 x 256 pixels, 0.98 x 0.98 mm in plane resolution, spacing between slices = 1 mm, slice thickness = 1 mm, inter-slice gap = 0). The anatomical slices covered the whole brain and were acquired parallel to the anterior-posterior commissure.
2.4. Data analysis in MRI
T2*BOLD volumes were preprocessed in SPM8 (Statistic Parametric Mapping, The Wellcome Trust Centre for Neuroimaging, University College London) [24] implemented in Matlab R2013b (TheMathWorks Inc., Massachussets, 2013) by applying realignment to correct motion artifacts and co-registration with structural image (T1-weighted volume). Five regions of interests (ROI) were selected on the subject’s anatomical image to identify changes in the magnetic signal intensity during the passage of the contrast agent. The ROIs were placed on the prefrontal cortex behind and in front of the FS and the right lateral prefrontal cortex as shown in (Fig. 1).
Fig. 1 Representation of the ROI positions on the prefrontal cortex behind of the frontal sinus (red), frontal sinus (gray), skull (yellow), skin (green) and on the right lateral prefrontal cortex (blue) of subject A in (a) axial and (b) sagittal view in a real space.
2.5. Procedure of the inflow of a dye in the DOT system
ICG is a non-toxic fluorescent dye that binds to serum proteins in plasma. This absorbs light in the near infrared spectrum with a maximum peak of absorption and emission at 805 nm in plasma solution [25]. Absorption changes due to ICG were measured by the CW-DOT device for brain blood inflow monitoring in subject A. The frames of the first 5 minutes were acquired before contrast agent injection to establish a pre-contrast baseline. Afterwards, a bolus of 12.5 mg of ICG diluted in 7.5 ml of saline was injected into the cubital vein of the right arm followed by a bolus of 5 ml of saline.
2.6. Repetitive transcranial magnetic stimulation procedures
Two stimulation probes were performed using a Magstim Transcranial Magnetic Stimulator 220 (Carmarthenshire, UK), in two different sessions, to measure hemodynamic changes evoked by rTMS on the prefrontal cortex of subject B.
As the aim of the present work is to study the influence of the FS on measures obtained by the CW-DOT system, the selected study areas were the medial prefrontal cortex (MPC) and the right lateral prefrontal cortex (RLPC). The RLPC has a smaller distance between skin and brain than the MPC allowing the comparison of changes for HbO and HbR in each study region.
A Brainsight neuro-navigational system (Rogue Resolutions Ltd, Cardiff, UK) was used to aid the precise placement of the TMS circular coil on the subject’s head using a previously acquired T1-weighted structural MRI image volume of the subject’s head. Prior to stimulation of the interest areas in the prefrontal cortex, the resting motor threshold (RMT) was determined in subject B in each session to apply an intensity output relative to the motor threshold on the prefrontal cortex in both the MPC and RLPC positions. RMT is defined as the minimum magnetic stimulation intensity applied to the motor cortex required to induce a reliable motor response such as that of an electromyogram (EMG) of the left first dorsal interosseus muscle [26]. In addition, a 50% frequency criterion to determine RMT was used, beginning at a clearly supra-threshold intensity. The RMT was defined when the EMG response showed an amplitude of ≥50 µV peak to peak [27].
The rTMS protocol at low frequency was performed by five trains of rTMS at 1 Hz, with a pulse length of 1 sec, 20 pulses per train and a 25 sec inter-train interval and at a 90° coil orientation to the MPC, in the first session. The same protocol was administered on RLPC in a second session (Fig. 2).
Fig. 2 Scheme of the repetitive transcranial magnetic stimulation (rTMS) protocol. Black blocks indicate trains of rTMS and violet blocks indicate inter-train intervals. The upper row depicts the time of each block during rTMS at low frequency (≥1 Hz).
The protocol of rTMS at low frequency used is a common protocol [28]. Although, low-frequencies of TMS have been associated with inhibitory effects on cortical physiology [29], showing better intracortical reproducibility than high-frequencies [30].
2.7. Optical data acquisition
The DYNOT 232 instrument (NIRx Medical Technologies. New York, USA) was used for monitoring the dual study on both subjects A & B. The system performs continuous-wave time multiplexed measurements using two frequency-encoded laser sources at 760nm and 830nm. Sampling rates of 1.8Hz (552 msec/volume) and 2.86Hz (350 msec/volume) were used for ICG-tracking and hemodynamic changes induced by rTMS monitoring, respectively. The NIR light travelled to and from the DOT device by optic fibers known as optodes. The optodes were arranged in a grid with a distance of one centimeter between them for the dual study.
The ICG-tracking study used 64 fiber optic probes all of which acted as detectors and 32 of them acted as sources (co-located) thereby providing 2048 optical channels. The grid was placed over the prefrontal cortex above the zygomatic arch until the Fz position referring to the EEG 10-20 System [31] (Fig. 3).
Fig. 3 (a) Localization of the study’s target volume, partially covering the frontal cortex including the frontal sinus. (b) Localizations of the rectangular grid containing the optical fibers (dots) on the boundary. Red dots correspond to source and all of them act as detectors.
The same optical instrument mentioned above measured hemodynamic changes induced by electric changes during rTMS probes. A co-located grid of 18 optic fibers was used, providing 324 optical channels. The grid of optical fibers was placed inside the circular TMS coil with a distance of 1 cm between optodes, defined from now on as TMS-DOT setup (Fig. 4).
Fig. 4 TMS-DOT setup. (a) Localizations of optical fibers inside the circular TMS coil to monitor the hemodynamic changes during rTMS probes. Optical fibers are co-located, all act as sources and as detectors providing 324 optical channels. (b) Position of TMS-DOT setup on a phantom.
2.8. Data quality
Optical signals may be contaminated by signals that are mixed during the acquisition of optical signals in a physical environment and to generate distortions to original signals affecting visualization and analysis [32]. Moreover, the optical signals propagating through the brain contain several spontaneous fluctuations originating from cardiac pulsation, respiration and change of blood pressure, which contaminate the signals measured by DOT, and which can induce spatial and time changes [33].
2.8.1. Filtering detector readings during the ICG-tracking
The absorption changes of the NIR light are accompanied by signals that are considered as noise during DOT measurements from the physical environment. A tool widely used in analysis and processing signal for noise reduction, data compression and peak detection is the discrete wavelet transform (DWT) algorithm [34]. As our interest was to detect the maximum peak corresponding to the ICG absorption in the prefrontal cortex of subject A, a DWT algorithm implemented in MATLAB that performs denoising of 1-D signals was applied to represent detector readings throughout the experimental period.
2.8.2. Filtering raw DOT data from rTMS-DOT recording
Physiological recordings throughout the experimental period during the rTMS probes were monitored in subject B. The subject wore sensors to measure cardiac cycle and breathing which were recorded continuously at 4 KHz using AD Instrument ML870/P PowerLab16/30 as the measurement system. The physiological data were used to remove physiological noise from raw DOT data using the dynamic retrospective filtering of physiological noise (DRIFTER) algorithm which applies an accurate dynamical tracking of the variations in the cardiac and respiratory frequencies [35].
2.9. Forward model and image reconstruction in rTMS probes
The positions of the fiber grid on subject B were defined on the FE-mesh, prior to forward modeling and image reconstruction. The use of a pre-calculated FE-mesh over a generic head model requires the translocation of the subject’s anatomy to the anatomic MR scans of the generic head model on which the FE-mesh is based. The spatial normalization SPM8 tool, which computes an affine and a non-linear transformation between two volumes was used [36]. Therefore, the position of the grid was interpolated between the coordinates marked by the Brainsight neuro-navigational system and assigned to the FE-mesh [35,37].
The BrainModeler tool (NIRx NAVI imaging) provides a library of sub-mesh based FE models [38], whose forward solutions of the photon diffusion equation (diffuse approximation DA) for all possible source and detector combinations on the sub-mesh’s boundary are pre-calculated. The sub-mesh that best approximated the area of the measurements according to the translocated positions of the fiber grid was selected. The sub-mesh contained 4921 nodes and 21144 tetrahedrons whose dimensions were 12.05 cm (width) x 8.85 cm (height) x 5.46 cm(thickness). Two positions on the prefrontal cortex were selected to compare hemodynamic changes during rTMS-DOT recordings and their configurations of the position of optic fibers for both MPC and RLPC positions are shown in (Fig. 5).
Fig. 5 Localizations of optical fibers (red dots) on the sub-mesh selected during rTMS monitoring. Each red dots correspond to co-located source and detectors pair on (a) the medial prefrontal cortex (MPC) and (b) on the right lateral side of the prefrontal cortex (RLPC) of subject B.
A weight matrix is obtained as a result of the relationship between the number of nodes of the sub-mesh (4921) and optical channels (324) measured by the combination of 18 sources and 18 detectors located on the head surface:
where,is the vector containing the measures for all source-detector combinations and, is the vector containing optical properties.
The normalized difference method [39] was used to reconstruct DOT volumes. The method relates, according to a perturbation approach [40], measured surface data with changes in interior optical properties of the medium used compared to a reference medium. The absorption changes at two wavelengths led to reconstructing images of relative HbO and HbR concentrations using extinction coefficients for both wavelengths [41].
The rebuilding of the DOT images requires inverting the weight matrix resulting in an ill-posed problem formulated as an inverse problem, owing to the fact that NIR light is highly attenuated with an increasing depth. The weight matrix is inverted by the use of a singular value decomposition (SVD) algorithm using minimum description length (MDL) criterion to select the number of singular values which explain the dimensionality of the matrix [35].
Finally, DOT images were reconstructed with a size of 64x64x64 voxels, in Analyze format for each hemoglobin state (HbO & HbR). The reconstructed DOT volumes were co-registered to the subject’s anatomy using the spatial normalization tool [36] to represent the brain activation on the subject’s anatomy.
3. Results
3.1. Changes in the magnetic susceptibility during the inflow of a contrast agent in an MRI device
Changes in the magnetic signal intensity were analyzed within the two intracerebral ROIs selected on the medial prefrontal cortex (behind the FS) and on the right lateral prefrontal cortex of subject A. Moreover, the extracerebral ROIs selected on the frontal sinus, skull and skin of subject A were analyzed to detect the arrival of the contrast agent to these regions. The magnetic signals are normalized with respect to the basal period, which corresponds to the first 5 sec before contrast agent injection. The curves are characterized by the arrival of the contrast agent with a sharp decline of the signal intensities. The signals are recovered after a peak signal as a result of washing out the contrast from the brain, as shown in (Fig. 6).
Fig. 6 Changes in the magnetic signal intensity within the selected ROIs (bottom right image). The abscissas axis represents the time in seconds and the ordinate axis corresponds to the magnetic signal intensity normalized to basal time. The red line represents the change in the signal intensity during the Gd inflow within the selected medial prefrontal cortex, behind the frontal sinus (red), frontal sinus (grey), skull (yellow), skin (green) ROIs of subject A. The blue line represents the change in the signal intensity during the Gd inflow within the selected right lateral prefrontal cortex ROI of subject A. Dashed line represents the end of the injection time.
The above technique for measuring cerebral inflow compares the signal intensity of voxels before and after the injection of a known contrast agent. The signal drop is directly related to the concentration of the contrast agent in the blood pool [42]. The results show that the max. signal peaks were reached at 18 sec for both selected intracerebral ROIs (RLPC & IMPC) including the basal time. The max. signal peaks were delayed in time for selected extracerebral ROIs such as the FS, skull and skin with respect to intracerebral ROIs of subject A.
3.2. Detector readings analysis during the inflow of a dye in a DOT system
DOT uses the multi-distance approach with the purpose of increasing spatial resolution and positional accuracy of optical brain imaging. This approach allows one to assume that the NIR light that is detected far away from the source has passed through deeper tissue layers, and the light that is detected close to the source has passed through superficial layers [6]. Therefore, signals measured from the first nearest neighbor to 10mm distance between sources and detectors (SD), contain information on extracerebral areas such as skull and skin tissue because they are less sensitive to deep tissues. While the signals measured from the third and fourth nearest neighbor to 30 and/or 40 mm distance between SD, contain components from brain tissue [18].
The detector readings to distances of 10, 20, 30 and 40 mm from a source placed on the prefrontal cortex were selected to measure the changes in the signal intensities during the ICG-tracking. Figure. 7 represents the position of a source and four selected detectors which covered the FS in the real space.
Fig. 7 (a) Localizations of a rectangular grid containing the optical fibers on the boundary of the prefrontal cortex in a real space. (b) Representation of the fiber’s grid. Red dots correspond to source and all of them act as detectors. The NIR light, which follows a banana path (arrows), is detected to a 10 mm (ch1), 20 mm (ch2), 30 mm (ch3) and 40 mm (ch4) distance from a source (S) over the prefrontal cortex (rectangles).
To measure absorption changes of the NIR light produced by ICG, the grid of the optodes was divided in three regions that cover: the right lateral prefrontal cortex (RLPC), superior medial prefrontal cortex (SMPC) and inferior medial prefrontal cortex (IMPC). It was decided to make the above divisions because the anatomical localization of the prefrontal cortex is behind the FS and the size of the FS is larger in medial regions than in the lateral regions.
The mean average of the detector readings during the ICG-tracking with a distance from a source of 10mm, 20mm, 30mm and 40mm for both wavelengths (760 & 830 nm) including the basal time (300 sec) were analyzed in the RLPC, where the scalp-brain distance was 8 mm in the real space as shown in (Fig. 8).
Fig. 8 (a) Localizations of a rectangular grid containing the optical fibers on the boundary of the right lateral prefrontal cortex whose scalp-brain distance was 8 mm (bottom right image). Representation of the time course of detector readings during the ICG-tracking with a distance from source of 10 mm, 20 mm, 30mm and 40 mm for both wavelengths (b) 760 nm and (c) 830 nm. The abscissas axis represents the experimental time in seconds and the ordinate axis corresponds to the changes in the NIR signal intensity normalized to basal time. Dashed lines represent the start and the end of the ICG injection period. Lower left graphics depict a zoom from the start of the injection (302 sec) to 350 sec.
The medial prefrontal cortex was divided in two regions; the SMPC and IMPC, where the scalp-brain distance was 9 mm and 16 mm respectively, to select the max. depth of the FS, which corresponds to the IMPC region. The time course of the mean average of the detector readings during the ICG-tracking with a distance from a source of 10mm, 20mm, 30mm and 40mm for both wavelengths (760 & 830 nm) including the basal time (300 sec), were analyzed in the SMPC as shown in (Fig. 9) and in the IMPC as shown in (Fig. 10).
Fig. 9 (a) Localizations of a rectangular grid containing the optical fibers on the boundary of the superior medial prefrontal cortex whose scalp-brain distance was 9 mm (bottom right image). Representation of the time course of detector readings during the ICG-tracking with a distance from source of 10 mm, 20 mm, 30mm and 40 mm for both wavelengths (b) 760 nm and (c) 830 nm. The abscissas axis represents the experimental time in seconds and the ordinate axis corresponds to the changes in the NIR signal intensity normalized to basal time. Dashed lines represent the start and the end of the ICG injection period. Lower left graphics depict a zoom from the start of the injection (302 sec) to 350 sec.
Fig. 10 (a) Localizations of a rectangular grid containing the optical fibers on the boundary of the inferior medial prefrontal cortex whose scalp-brain distance was 16 mm (bottom right image). Representation of the time course of detector readings during the ICG-tracking with a distance from source of 10 mm, 20 mm, 30mm and 40 mm for both wavelengths (b) 760 nm and (c) 830 nm. The abscissas axis represents the experimental time in seconds and the ordinate axis corresponds to the changes in the NIR signal intensity normalized to basal time. Dashed lines represent the start and the end of the ICG injection period. Lower left graphics depict a zoom from the start of the injection (302 sec) to 350 sec.
The results from raw DOT data show that for both wavelengths in the RLPC, the max. signal peaks were reached in the second 325.1 for the detectors placed at a distance from a source of 30mm & 40 mm. While, the detectors placed at shorter distances from a source (10mm & 20mm) show that the max. signal peaks were delayed in time (327.9 sec & 325.7 sec, respectively).
The detector readings for both wavelengths in the medial prefrontal cortex (SMPC & IMPC) show that, the max. signal peaks were reached in the second 325.1 for the detectors placed at a distance of 40mm from source. While, the detectors placed at smaller distances from a source (10mm) show that the max. signal peaks were delayed in time (328.4 sec).
The detector readings indicate that the measurements at a short distance between one source and one detector are a few seconds later than those at greater distances, as has been previously described by other authors, even using variable optical technology. For example, the capability of NIRS to noninvasively monitor cerebral perfusion in the human adult based on an ICG-bolus tracking was demonstrated using a frequency domain technique to measure the attenuation light changes. Here, the arrival of the bolus signal appears to be delayed in the upper layer (skin and skull) compared to that in the lower layer (brain) [18]. Using a time-resolved instrument to record distributions of times of flight of photons at source–detector separations of 1.5, 2, 2.5, and 3 cm before and after application of a bolus of ICG showed that the dynamics of the upper layers were slower. In this study, an initial rise in the change in the absorption coefficient occurred at a later time followed by a slow washout period. The different dynamics observed were interpreted to be characteristic of intracerebral and extracerebral compartments [17].Finally, the feasibility of separating intra- and extracerebral tissue according to the arrival of the ICG bolus in motor area was investigated using a continuous wave instrument. Where a high lateral and good depth resolution was demonstrated which allowed the separation of intra- and extracerebral tissue [19].Therefore, at longer distances between source-detector, the signals provide information from deep tissues.
The end of the injection is considered as t = 0 as the reference to compare the measurements of both CW-DOT and MRI devices. The volumes and injection duration for each contrast were different due to the properties of each contrast. Taking into account the above, the signals measured from MRI and CW-DOT were compared for the intracerebral regions such as the cerebral cortex behind the FS (IMPC) and in the right lateral prefrontal cortex (RLPC) of subject A. The results show that: for optical measures from detectors placed at greater distances from a source (40mm) and magnetic susceptibility measures from the selected intracerebral ROIs, the max. signal peaks were reached in the 11th-12th sec after the injection of both blood markers (ICG and Gd) as shown in (Fig. 11).
Fig. 11 Representation of the marked cerebral blood dynamic measured by MRI and CW-DOT devices from t = 0 (the end of the injection). The max. signal peaks are shown for intracerebral areas behind the FS measured by a CW-DOT device (thick red line) and MRI device (fine red line). The max. signal peaks on the right lateral prefrontal cortex measured by a CW-DOT device (thick blue line) and MRI device (fine blue line). The abscissas axis represents the experimental time in seconds and the ordinate axis corresponds to the changes in the normalized signal intensities. The time of arrival for marked cerebral blood in intracerebral areas (dashed line).
Considering the end of the injection as t = 0 as the reference, the times of the max. signal peaks, in other words, the max. absorption of NIR light by ICG, may be compared for optical measurements at different distances between one source and one detector (optode separation) placed on the region of interest (ROIs): the RLPC, the SMPC and the IMPC. The differences in the times for the maximum peaks of absorption ICG were tested using 3 ROIs x 4 optode separation (10, 20, 30, 40 mm). ANOVA assumptions were tested using Levene's test for homogeneity of variances and Kolmogorov–Smirnov's test for normality. There was no significant main effect of ROI (p = 0.156). There was no significant ROI x optode separation interaction (p = 0.969). There was a significant main effect of optode separation (p<.001), and Dunnett's post hoc test showed a significant difference (p<.001) between the mean of the group of 40 mm and the mean of any other group of 10, 20 and 30 mm. Analysis was performed according to ANOVA’s comparison using a GraphPad Statistics Guide (http://www.graphpad.com/.htm) and Matlab R2013b.
Moreover, the results show that in three ROIs, the optical measurements for the detectors placed 40 mm from a source, the max. peaks are reached in the 11th −12th second. In addition, as the distance between detectors-sources on the head’s surface increases, the max. signal peaks appear a few seconds delayed in time. The same results can be seen in magnetic susceptibility measurements on the IMPC, where the max. signal peak is reached in the 11th −12th second corresponding to the brain ROI. While the max. signal peak for extracerebral regions such as the FS, skull and skin appears a few seconds delayed in the time (Fig. 12).
Fig. 12 (a) The histogram depicts the times for the max. peaks of absorption ICG for detectors placed 10, 20, 30 and 40 mm from one source on the right lateral prefrontal cortex (RLPC), superior medial prefrontal cortex (SMPC) and inferior medial prefrontal cortex (IMPC) of subject A (* p-value≤0.001; NS: no significative). (b) The histogram represents the times for the max. peaks of Gd within the selected skin, skull, frontal sinus and brain ROIs on the inferior medial prefrontal cortex of subject A.
3.3. rTMS-DOT recordings on the medial prefrontal cortex (MPC)
Absorption changes at two wavelengths led to reconstructed three-dimensional images of relative changes for HbO, HbR and HbT using extinction coefficients of HbO and HbR for both wavelengths [41]. HbT represents the hemodynamic changes produced by rTMS at 1 Hz on the MPC of subject B.
The TMS-DOT setup was placed over the MPC to measure hemodynamic changes produced by rTMS across the FS. These changes are represented in three-dimensional DOT images co-registered to the subject’s anatomy from a generic pre-calculated FE-mesh. The sagittal and axial views of a slice of the subject’s anatomy corresponding to one DOT volume in time are shown in (Fig. 13).
Fig. 13 Representation of (a) sagittal and (b) axial view of a reconstructed DOT volume co-registered to the subject’s anatomy from a pre-calculated FE-mesh. The TMS-DOT setup (red line) was placed over the medial prefrontal cortex to measure across the frontal sinus. The orange line depicts the distance from the cerebral cortex to the scalp in real space. The color bar indicates changes in HbT (10−5) within a train of rTMS at 1 Hz.
Reconstructed DOT volumes show an increase of HbT in the cerebral cortex target behind the FS during a rTMS session at a low frequency. The TMS-DOT setup detected cerebral activation across the FS, where the scalp-brain distance was 26.4 mm in real space. It was not possible to detect hemodynamic changes beyond the measurement zone due to the position and the number of optical fibers.
The multi-distance approach used by DOT allows the detection of hemodynamic changes in both extracerebral and intracerebral areas as shown in (Fig. 13). As the interest here was only the cerebral cortex, the time courses during the whole experiment for each hemoglobin state HbO & HbR were analyzed using an ROI analysis.
Changes in HbR and HbO during 20 seconds of rTMS (blue bars) for each rTMS block within the ROI located on the MPC behind the FS are shown in (Fig. 14).
Fig. 14 (a) ROI position on a slice from a reconstructed HbT to measure the hemodynamic changes behind the frontal sinus. (b) Representation of the time course of HbO (red line) and HbR (blue line) within the ROI selected during rTMS at 1 Hz. Blue bars represent the duration of each rTMS block (20 sec). The abscissas axis represents the time in seconds and the ordinate axis corresponds to micromolar concentration (10−6).
Time series from the selected ROI show an increase for both hemoglobin states (HbO & HbR) during the rTMS blocks. During interstimuli period (25 sec), the signals decrease near to basal time until the next rTMS block.
The time series clearly represent the hemodynamic changes according to the stimulation block. Therefore, both the reconstructed images and time series for both hemoglobin states show the capacity of a CW-DOT system to measure hemodynamic changes in the cerebral cortex across the FS using rTMS at low frequency.
3.4. rTMS-DOT recordings on the right lateral prefrontal cortex (RLPC)
Moreover, the TMS-DOT setup was placed on the RLPC of subject B to measure hemodynamic changes produced by rTMS at 1 Hz. In addition, the results can be compared with the results generated from the measurements on the MPC. Hemodynamic changes are represented in three-dimensional DOT images co-registered to the subject’s anatomy from a generic pre-calculated FE-mesh. The sagittal and axial views of a slice of the subject’s anatomy corresponding to one DOT volume in time are shown in (Fig. 15).
Fig. 15 Representation of (a) sagittal and (b) axial view of a reconstructed DOT volume co-registered to the subject’s anatomy from a pre-calculated FE-mesh. The TMS-DOT setup (red line) was placed over the right lateral prefrontal cortex (RLPC). The orange line depicts the distance from the cerebral cortex to the scalp in real space. The color bar indicates changes in HbT (10−4) within a train of rTMS at 1 Hz.
Scalp-brain distance was 13.5 mm in real space, where the light could penetrate a greater distance using the same TMS-DOT setup. For example, measurements on the RLPC show higher loci activations than measurements on the MPC from reconstructed DOT volumes.
Time courses during the whole experiment for each hemoglobin state (HbO & HbR) were analyzed on the RLPC using an ROI analysis. Changes for HbO and HbR during 20 seconds of rTMS (blue bars) for each rTMS block within the ROI located in the interest area are shown in (Fig. 16).
Fig. 16 (a) ROI position on a slice from reconstructed HbT to measure the hemodynamic changes on the right lateral prefrontal cortex (RLPC). (b) Representation of the time course of HbO (red line) and HbR (blue line) within the selected ROI during the rTMS at 1 Hz. Blue bars represent the duration of each rTMS block (20 sec). The abscissas axis represents the time in seconds and the ordinate axis corresponds to micromolar concentration (10−7).
Time series from the selected ROI show an increase for both hemoglobin states (HbO & HbR) during the rTMS blocks at low frequency. The present result shows that the first rTMS generated an increase of both hemoglobin states, unlike the rest of the rTMS blocks that are less variable. These abrupt hemodynamic changes could be due to a neuronal adaptation once an activation threshold is exceeded.
4. Discussion
The results show the capability and feasibility of a CW-DOT system to measure hemodynamic changes across the FS until the cerebral cortex target. The influence of the FS on the sensitivity of NIR light is variable, depending on its depth, thickness of the skull and optical properties of the scalp and skull [12]. Furthermore, the size and depth of the FS varies between individuals [13].
4.1. Inflow studies
Previous research has demonstrated that a DOT device has sufficient depth and lateral resolution to be used in cerebral blood inflow monitoring, making it possible to separate intra- and extra-cerebral tissues using ICG as a blood dye [18,19]. Against this background, it was decided to use a CW-DOT device to monitor marked cerebral blood inflow in the prefrontal cortex behind the FS.
The detector readings in the three regions show that max. signal peaks are delayed a few seconds when the detectors are placed at a short distance from a source compared to the detectors placed at a longer distance from a source when the ICG passed across the monitored regions. These results show that the optical channels contain intracerebral information at longer distances between SD while at shorter distances between SD they contain extracerebral information, as reported by other authors [18,19,43]. Given that most of the photons inside the head follow a banana shaped path [15,44], it is possible to estimate the relationship between the distance from one source to one detector placed on the head surface and the depth of penetration of the photons inside the head. These estimates suggest that the photons can reach approx. 2 cm of depth for a distance of 3 cm between a source and a detector [13,15].
The results of the present work show that for both ICG and Gd tracking in the prefrontal cortex behind the frontal sinus, the max. signal peaks were reached in the 11th −12th second, after injecting subject A. Therefore, the absorption changes occurring at greater distances between sources-detectors (40mm) on the prefrontal cortex are sampled from brain tissue, as happens in MRI. The results suggest that the dynamic of the marked cerebral blood inflow measured by MRI and CW-DOT devices indicate that the cerebral blood flow appears in the cerebral cortex and then spreads towards extracerebral areas. (Fig. 17).
Fig. 17 Scheme of the marked cerebral blood dynamic measured by MRI and a DOT device in a real space. A few seconds after injection, the marked blood arrives in the cortex (red) followed by its arrival in the extracerebral region (violet) seconds later, due to a washout from the brain.
In summary, both ICG and Gd tracking on the prefrontal cortex of subject A measured by both neuroimaging techniques show that it is possible to measure the NIR light absorption changes in the cerebral cortex target behind the FS using CW-DOT systems. Future studies about the dynamic for both ICG tracking could be used to diagnose or as a functional probe in the cerebral tumor as has been shown in the breast [45].
4.2. Repetitive transcranial magnetic stimulation studies
The ICG-tracking using a CW-DOT system provides extra and intracerebral information of the prefrontal cortex, but it is not completely clear which is why we decided to perform the rTMS study, which is more decisive than the inflow study. TMS is a versatile and noninvasive technique, which applies a discharge of an electric current through a wire coil placed near the head surface generating a magnetic flux that induces a weak current in the cortex. Nevertheless, this weak current is sufficient to generate action potentials according to the coil configuration and placement [22].
The results from the rTMS-DOT recordings at low frequency presented here show an increase in both hemoglobin states on both the MPC and the RLPC positions in subject B. The time series show a sustained HbO increase during rTMS of 20 sec for both positions. These activation patterns were detected in studies with rTMS at a low frequency on the motor cortex [46] and on the ipsi and contralateral prefrontal cortex [47], all of which were simultaneously monitored by optical imaging. In addition, the time series show a sustained increase of the HbR during the rTMS (20 sec) in both the MPC and the RLPC positions. These activation patterns have been previously described on the ipsi and contralateral prefrontal cortex during rTMS at 1 Hz, monitored by optical imaging [48].
The hemodynamic changes measured by the rTMS-DOT recordings at a low frequency could be associated with a vasodilatation because a long stimulation period (20 sec) requires oxygen contribution and the release of waste products generated by neuronal activity [49].
Previous studies have reported clear differences in the TMS effect based on intensity, orientation of the coil and stimulation area [50]. Thus, different intensities of rTMS elicit different patterns of hemodynamic response, even at the same stimulation frequency [50–52].
Besides the discrepancies, the present study in subject B shows the reliability and capability of CW-DOT for noninvasive measurement of cerebral hemodynamic changes across the FS during the administration of rTMS at low frequency. Furthermore, the tomography approach and reconstruction algorithms provide three-dimensional hemodynamic images during the stimulation period. With the method proposed in the present work based on TMS/DOT setup it is possible to generate a controlled cerebral response in the prefrontal cortex which can show deactivations during a task [53].
5. Conclusions
The FS is a structure occupying the region between the skull and the grey matter. The present dual study on the prefrontal cortex in two healthy subjects demonstrates the reliability of NIR light reaching the cerebral cortex target behind the FS using a CW-DOT device which also offers a multi-distance approach. There are some discrepancies in both studies due to lack of the knowledge and the fact that the interpretation of the results is not entirely clear. This lack of clarity arises from the fact that the study area is located far from the scalp and is comprised of both hemispheres that are curved-shape and extravascular structures. The only certain conclusion of the present study is that CW-DOT devices can measure hemodynamic changes in the cerebral cortex target across the FS. These measurements were performed in two healthy subjects with a known FS size. Nevertheless, the present work does not address the variability of the FS size between people and this should be studied in future works.
Funding
The Cooperation Program Interreg MAC (Madeira-Azores-Canarias) 2014-2020; the European Regional Development Fund (ERDF). (MAC/1.1b/098).
Acknowledgments
We would like to thank our volunteers for their participation in this study. We also wish to thank Jose Maria Perez González for his assistance with data acquisition and Patrick Dennis for his help in revising the English language in the manuscript. We would also like to acknowledge the support of the Servicio de Resonancia Magnética para Investigaciones Biomédicas de la Universidad de La Laguna.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. M. Ferrari and V. Quaresima, “A brief review on the history of human functional near-infrared spectroscopy (fNIRS) development and fields of application,” Neuroimage 63(2), 921–935 (2012). [CrossRef] [PubMed]
2. D. A. Boas, C. E. Elwell, M. Ferrari, and G. Taga, “Twenty years of functional near-infrared spectroscopy: introduction for the special issue,” Neuroimage 85, 1-15 (2014). [CrossRef]
3. A. Bluestone, G. Abdoulaev, C. Schmitz, R. Barbour, and A. Hielscher, “Three-dimensional optical tomography of hemodynamics in the human head,” Opt. Express 9(6), 272–286 (2001). [CrossRef] [PubMed]
4. D. T. Delpy, M. Cope, P. van der Zee, S. Arridge, S. Wray, and J. Wyatt, “Estimation of optical pathlength through tissue from direct time of flight measurement,” Phys. Med. Biol. 33(12), 1433–1442 (1988). [CrossRef] [PubMed]
5. A. Sassaroli and S. Fantini, “Comment on the modified Beer-Lambert law for scattering media,” Phys. Med. Biol. 49(14), N255–N257 (2004). [CrossRef] [PubMed]
6. R. L. Barbour, H. L. Graber, Y. Pei, S. Zhong, and C. H. Schmitz, “Optical tomographic imaging of dynamic features of dense-scattering media,” J. Opt. Soc. Am. A 18(12), 3018–3036 (2001). [CrossRef] [PubMed]
7. E. Okada, D. Yamamoto, N. Kiryu, A. Katagiri, N. Yokose, T. Awano, K. Igarashi, S. Nakamura, T. Hoshino, and Y. Murata, “Theoretical and experimental investigation of the influence of frontal sinus on the sensitivity of the NIRS signal in the adult head,” in Oxygen Transport to Tissue XXXI (Springer, 2010), pp. 231–236.
8. E. Okada, M. Firbank, M. Schweiger, S. R. Arridge, M. Cope, and D. T. Delpy, “Theoretical and experimental investigation of near-infrared light propagation in a model of the adult head,” Appl. Opt. 36(1), 21–31 (1997). [CrossRef] [PubMed]
9. D. Boas, J. Culver, J. Stott, and A. Dunn, “Three dimensional Monte Carlo code for photon migration through complex heterogeneous media including the adult human head,” Opt. Express 10(3), 159–170 (2002). [CrossRef] [PubMed]
10. Y. Fukui, Y. Ajichi, and E. Okada, “Monte Carlo prediction of near-infrared light propagation in realistic adult and neonatal head models,” Appl. Opt. 42(16), 2881–2887 (2003). [CrossRef] [PubMed]
11. E. Okada and D. T. Delpy, “Near-infrared light propagation in an adult head model. I. Modeling of low-level scattering in the cerebrospinal fluid layer,” Appl. Opt. 42(16), 2906–2914 (2003). [CrossRef] [PubMed]
12. K. Kurihara, H. Kawaguchi, T. Obata, H. Ito, K. Sakatani, and E. Okada, “The influence of frontal sinus in brain activation measurements by near-infrared spectroscopy analyzed by realistic head models,” Biomed. Opt. Express 3(9), 2121–2130 (2012). [CrossRef] [PubMed]
13. F. B. Haeussinger, S. Heinzel, T. Hahn, M. Schecklmann, A.-C. Ehlis, and A. J. Fallgatter, “Simulation of near-infrared light absorption considering individual head and prefrontal cortex anatomy: implications for optical neuroimaging,” PLoS One 6(10), e26377 (2011). [CrossRef] [PubMed]
14. S. Pirner, K. Tingelhoff, I. Wagner, R. Westphal, M. Rilk, F. M. Wahl, F. Bootz, and K. W. G. Eichhorn, “CT-based manual segmentation and evaluation of paranasal sinuses,” Eur. Arch. Otorhinolaryngol. 266(4), 507–518 (2009). [CrossRef] [PubMed]
15. X. Cui, S. Bray, D. M. Bryant, G. H. Glover, and A. L. Reiss, “A quantitative comparison of NIRS and fMRI across multiple cognitive tasks,” Neuroimage 54(4), 2808–2821 (2011). [CrossRef] [PubMed]
16. M. Wolf, O. Weber, M. Keel, X. Golay, M. Scheidegger, H. U. Bucher, S. Kollias, P. Boesiger, and O. Baenziger, “Comparison of cerebral blood volume measured by near infrared spectroscopy and contrast enhanced magnetic resonance imaging,” in Oxygen Transport to Tissue XXI (Springer, 1999), pp. 767–773.
17. A. Liebert, H. Wabnitz, J. Steinbrink, H. Obrig, M. Möller, R. Macdonald, A. Villringer, and H. Rinneberg, “Time-resolved multidistance near-infrared spectroscopy of the adult head: intracerebral and extracerebral absorption changes from moments of distribution of times of flight of photons,” Appl. Opt. 43(15), 3037–3047 (2004). [CrossRef] [PubMed]
18. M. Kohl-Bareis, H. Obrig, J. Steinbrink, J. Malak, K. Uludag, and A. Villringer, “Noninvasive monitoring of cerebral blood flow by a dye bolus method: separation of brain from skin and skull signals,” J. Biomed. Opt. 7(3), 464–470 (2002). [CrossRef] [PubMed]
19. C. Habermehl, C. H. Schmitz, and J. Steinbrink, “Contrast enhanced high-resolution diffuse optical tomography of the human brain using ICG,” Opt. Express 19(19), 18636–18644 (2011). [CrossRef] [PubMed]
20. E. A. Knopp, S. Cha, G. Johnson, A. Mazumdar, J. G. Golfinos, D. Zagzag, D. C. Miller, P. J. Kelly, and I. I. Kricheff, “Glial neoplasms: dynamic contrast-enhanced T2*-weighted MR imaging,” Radiology 211(3), 791–798 (1999). [CrossRef] [PubMed]
21. O. Yamashita, T. Shimokawa, R. Aisu, T. Amita, Y. Inoue, and M. A. Sato, “Multi-subject and multi-task experimental validation of the hierarchical Bayesian diffuse optical tomography algorithm,” Neuroimage 135, 287–299 (2016). [CrossRef] [PubMed]
22. M. Hallett, “Transcranial magnetic stimulation: a primer,” Neuron 55(2), 187–199 (2007). [CrossRef] [PubMed]
23. M. Essig, M. S. Shiroishi, T. B. Nguyen, M. Saake, J. M. Provenzale, D. Enterline, N. Anzalone, A. Dörfler, A. Rovira, M. Wintermark, and M. Law, “Perfusion MRI: the five most frequently asked technical questions,” AJR Am. J. Roentgenol. 200(1), 24–34 (2013). [CrossRef] [PubMed]
24. K. J. Friston, A. P. Holmes, K. J. Worsley, J. Poline, C. D. Frith, and R. S. J. Frackowiak, “Statistical parametric maps in functional imaging: a general linear approach,” Hum. Brain Mapp. 2(4), 189–210 (1994). [CrossRef]
25. M. L. Landsman, G. Kwant, G. A. Mook, and W. G. Zijlstra, “Light-absorbing properties, stability, and spectral stabilization of indocyanine green,” J. Appl. Physiol. 40(4), 575–583 (1976). [CrossRef] [PubMed]
26. J. C. Rothwell, M. Hallett, A. Berardelli, A. Eisen, P. Rossini, W. Paulus, and The International Federation of Clinical Neurophysiology, “Magnetic stimulation: motor evoked potentials,” Electroencephalogr. Clin. Neurophysiol. Suppl. 52, 97–103 (1999). [PubMed]
27. G. G. Westin, B. D. Bassi, S. H. Lisanby, B. Luber, and N. Y. S. P. Institute, “Determination of motor threshold using visual observation overestimates transcranial magnetic stimulation dosage: safety implications,” Clin. Neurophysiol. 125(1), 142–147 (2014). [CrossRef] [PubMed]
28. P. M. Rossini, A. T. Barker, A. Berardelli, M. D. Caramia, G. Caruso, R. Q. Cracco, M. R. Dimitrijević, M. Hallett, Y. Katayama, C. H. Lücking, A. L. Maertens de Noordhout, C. D. Marsden, N. M. F. Murray, J. C. Rothwell, M. Swash, and C. Tomberg, “Non-invasive electrical and magnetic stimulation of the brain, spinal cord and roots: basic principles and procedures for routine clinical application. Report of an IFCN committee,” Electroencephalogr. Clin. Neurophysiol. 91(2), 79–92 (1994). [CrossRef] [PubMed]
29. R. Chen, J. Classen, C. Gerloff, P. Celnik, E. M. Wassermann, M. Hallett, and L. G. Cohen, “Depression of motor cortex excitability by low-frequency transcranial magnetic stimulation,” Neurology 48(5), 1398–1403 (1997). [CrossRef] [PubMed]
30. F. Maeda, M. Gangitano, M. Thall, and A. Pascual-Leone, “Inter- and intra-individual variability of paired-pulse curves with transcranial magnetic stimulation (TMS),” Clin. Neurophysiol. 113(3), 376–382 (2002). [CrossRef] [PubMed]
31. G. H. Klem, H. O. Lüders, H. H. Jasper, C. Elger, and The International Federation of Clinical Neurophysiology, “The ten-twenty electrode system of the International Federation,” Electroencephalogr. Clin. Neurophysiol. Suppl. 52(3), 3–6 (1999). [PubMed]
32. V. Strela, J. Portilla, and E. P. Simoncelli, “Image denoising using a local Gaussian scale mixture model in the wavelet domain,” in Wavelet Applications in Signal and Image Processing VIII (International Society for Optics and Photonics, 2000), 4119, pp. 363–372.
33. F. Tian, H. Niu, B. Khan, G. Alexandrakis, K. Behbehani, and H. Liu, “Enhanced functional brain imaging by using adaptive filtering and a depth compensation algorithm in diffuse optical tomography,” IEEE Trans. Med. Imaging 30(6), 1239–1251 (2011). [CrossRef] [PubMed]
34. S. G. Mallat, “A Theory for Multiresolution Signal Decomposition: The Wavelet Representation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(7), 674–693 (1989). [CrossRef]
35. E. Hernandez-Martin, F. Marcano, O. Casanova, C. Modroño, J. Plata-Bello, and J. L. González-Mora, “Comparing diffuse optical tomography and functional magnetic resonance imaging signals during a cognitive task: pilot study,” Neurophotonics 4, 015003 (2017). [CrossRef]
36. K. Friston, J. Ashburner, C. D. Frith, J. Poline, J. D. Heather, and R. S. J. Frackowiak, “Spatial registration and normalization of images,” Hum. Brain Mapp. 3(3), 165–189 (1995). [CrossRef]
37. C. Habermehl, S. Holtze, J. Steinbrink, S. P. Koch, H. Obrig, J. Mehnert, and C. H. Schmitz, “Somatosensory activation of two fingers can be discriminated with ultrahigh-density diffuse optical tomography,” Neuroimage 59(4), 3201–3211 (2012). [CrossRef] [PubMed]
38. S. R. Arridge, M. Schweiger, M. Hiraoka, and D. T. Delpy, “A finite element approach for modeling photon transport in tissue,” Med. Phys. 20(2 Pt 1), 299–309 (1993). [CrossRef] [PubMed]
39. Y. Pei, H. L. Graber, and R. L. Barbour, “Influence of systematic errors in reference states on image quality and on stability of derived information for DC optical imaging,” Appl. Opt. 40(31), 5755–5769 (2001). [CrossRef] [PubMed]
40. K. D. Paulsen and H. Jiang, “Spatially varying optical property reconstruction using a finite element diffusion equation approximation,” Med. Phys. 22(6), 691–701 (1995). [CrossRef] [PubMed]
41. D. A. Boas and A. M. Dale, “Simulation study of magnetic resonance imaging-guided cortically constrained diffuse optical tomography of human brain function,” Appl. Opt. 44(10), 1957–1968 (2005). [CrossRef] [PubMed]
42. A. Majos, P. Bogorodzki, E. Piątkowska-Janko, T. Wolak, R. Kurjata, and L. Stefańczyk, “Functional imaging with MR T1 contrast: a feasibility study with blood-pool contrast agent,” Eur. Radiol. 19(4), 898–903 (2009). [CrossRef] [PubMed]
43. A. Liebert, H. Wabnitz, H. Obrig, R. Erdmann, M. Möller, R. Macdonald, H. Rinneberg, A. Villringer, and J. Steinbrink, “Non-invasive detection of fluorescence from exogenous chromophores in the adult human brain,” Neuroimage 31(2), 600–608 (2006). [CrossRef] [PubMed]
44. P. Van der Zee and D. T. Delpy, “Simulation of the point spread function for light in tissue by a Monte Carlo method,” in Oxygen Transport to Tissue IX (Springer, 1987), pp. 179–191.
45. V. Ntziachristos, A. G. Yodh, M. Schnall, and B. Chance, “Concurrent MRI and diffuse optical tomography of breast after indocyanine green enhancement,” Proc. Natl. Acad. Sci. U.S.A. 97(6), 2767–2772 (2000). [CrossRef] [PubMed]
46. R. C. Mesquita, O. K. Faseyitan, P. E. Turkeltaub, E. M. Buckley, A. Thomas, M. N. Kim, T. Durduran, J. H. Greenberg, J. A. Detre, A. G. Yodh, and R. H. Hamilton, “Blood flow and oxygenation changes due to low-frequency repetitive transcranial magnetic stimulation of the cerebral cortex,” J. Biomed. Opt. 18(6), 067006 (2013). [CrossRef] [PubMed]
47. N. Hanaoka, Y. Aoyama, M. Kameyama, M. Fukuda, and M. Mikuni, “Deactivation and activation of left frontal lobe during and after low-frequency repetitive transcranial magnetic stimulation over right prefrontal cortex: a near-infrared spectroscopy study,” Neurosci. Lett. 414(2), 99–104 (2007). [CrossRef] [PubMed]
48. F. A. Kozel, F. Tian, S. Dhamne, P. E. Croarkin, S. M. McClintock, A. Elliott, K. S. Mapes, M. M. Husain, and H. Liu, “Using simultaneous repetitive Transcranial Magnetic Stimulation/functional Near Infrared Spectroscopy (rTMS/fNIRS) to measure brain activation and connectivity,” Neuroimage 47(4), 1177–1184 (2009). [CrossRef] [PubMed]
49. A. Devor, P. Tian, N. Nishimura, I. C. Teng, E. M. C. Hillman, S. N. Narayanan, I. Ulbert, D. A. Boas, D. Kleinfeld, and A. M. Dale, “Suppressed neuronal activity and concurrent arteriolar vasoconstriction may explain negative blood oxygenation level-dependent signal,” J. Neurosci. 27(16), 4452–4459 (2007). [CrossRef] [PubMed]
50. R. H. Thomson, T. J. Cleve, N. W. Bailey, N. C. Rogasch, J. J. Maller, Z. J. Daskalakis, and P. B. Fitzgerald, “Blood oxygenation changes modulated by coil orientation during prefrontal transcranial magnetic stimulation,” Brain Stimul. 6(4), 576–581 (2013). [CrossRef] [PubMed]
51. R. H. Thomson, J. J. Maller, Z. J. Daskalakis, and P. B. Fitzgerald, “Blood oxygenation changes resulting from trains of low frequency transcranial magnetic stimulation,” Cortex 48(4), 487–491 (2012). [CrossRef] [PubMed]
52. Z. Nahas, M. Lomarev, D. R. Roberts, A. Shastri, J. P. Lorberbaum, C. Teneback, K. McConnell, D. J. Vincent, X. Li, M. S. George, and D. E. Bohning, “Unilateral left prefrontal transcranial magnetic stimulation (TMS) produces intensity-dependent bilateral effects as measured by interleaved BOLD fMRI,” Biol. Psychiatry 50(9), 712–720 (2001). [CrossRef] [PubMed]
53. F. S. Racz, P. Mukli, Z. Nagy, and A. Eke, “Increased prefrontal cortex connectivity during cognitive challenge assessed by fNIRS imaging,” Biomed. Opt. Express 8(8), 3842–3855 (2017). [CrossRef] [PubMed]
