Direct approach to compute Jacobians for diffuse optical tomography using perturbation Monte Carlo-based photon "replay"

Maternal prenatal depressive symptoms and child brain responses to affective touch at two years of age

BACKGROUND

Touch is an essential form of mother-child interaction, instigating better social bonding and emotional stability.

METHODS

We used diffuse optical tomography to explore the relationship between total haemoglobin (HbT) responses to affective touch in the child's brain at two years of age and maternal self-reported prenatal depressive symptoms (EPDS). Affective touch was implemented via slow brushing of the child's right forearm at 3 cm/s and non-affective touch via fast brushing at 30 cm/s and HbT responses were recorded on the left hemisphere.

RESULTS

We discovered a cluster in the postcentral gyrus exhibiting a negative correlation (Pearson's r = -0.84, p = 0.015 corrected for multiple comparisons) between child HbT response to affective touch and EPDS at gestational week 34. Based on region of interest (ROI) analysis, we found negative correlations between child responses to affective touch and maternal prenatal EPDS at gestational week 14 in the precentral gyrus, Rolandic operculum and secondary somatosensory cortex. The responses to non-affective touch did not correlate with EPDS in these regions.

LIMITATIONS

The number of mother-child dyads was 16. However, by utilising high-density optode arrangements, individualised anatomical models, and video and accelerometry to monitor movement, we were able to minimize methodological sources of variability in the data.

CONCLUSIONS

The results show that maternal depressive symptoms during pregnancy may be associated with reduced child responses to affective touch in the temporoparietal cortex. Responses to affective touch may be considered as potential biomarkers for psychosocial development in children. Early identification of and intervention in maternal depression may be important already during early pregnancy.

Computationally-efficient linear scheme for overlap time-gating spatial frequency domain diffuse optical tomography using an analytical diffusion model

Time-domain (TD) spatial frequency domain (SFD) diffuse optical tomography (DOT) potentially enables laminar tomography of both the absorption and scattering coefficients. Its full time-resolved-data scheme is expected to enhance performances of the image reconstruction but poses heavy computational costs and also susceptible signal-to-noise ratio (SNR) limits, as compared to the featured-data one. We herein propose a computationally-efficient linear scheme of TD-SFD-DOT, where an analytical solution to the TD phasor diffusion equation for semi-infinite geometry is derived and used to formulate the Jacobian matrices with regard to overlap time-gating data of the time-resolved measurement for improved SNR and reduced redundancy. For better contrasting the absorption and scattering and widely adapted to practically-available resources, we develop an algebraic-reconstruction-technique-based two-step linear inversion procedure with support of a balanced memory-speed strategy and multi-core parallel computation. Both simulations and phantom experiments are performed to validate the effectiveness of the proposed TD-SFD-DOT method and show an achieved tomographic reconstruction at a relative depth resolution of ∼4 mm.

Enhanced model iteration algorithm with graph neural network for diffuse optical tomography

Diffuse optical tomography (DOT) employs near-infrared light to reveal the optical parameters of biological tissues. Due to the strong scattering of photons in tissues and the limited surface measurements, DOT reconstruction is severely ill-posed. The Levenberg-Marquardt (LM) is a popular iteration method for DOT, however, it is computationally expensive and its reconstruction accuracy needs improvement. In this study, we propose a neural model based iteration algorithm which combines the graph neural network with Levenberg-Marquardt (GNNLM), which utilizes a graph data structure to represent the finite element mesh. In order to verify the performance of the graph neural network, two GNN variants, namely graph convolutional neural network (GCN) and graph attention neural network (GAT) were employed in the experiments. The results showed that GCNLM performs best in the simulation experiments within the training data distribution. However, GATLM exhibits superior performance in the simulation experiments outside the training data distribution and real experiments with breast-like phantoms. It demonstrated that the GATLM trained with simulation data can generalize well to situations outside the training data distribution without transfer training. This offers the possibility to provide more accurate absorption coefficient distributions in clinical practice.

Wide-field illumination diffuse optical tomography within a framework of single-pixel time-domain spatial frequency domain imaging

We present a wide-field illumination time-domain (TD) diffusion optical tomography (DOT) for three-dimensional (3-D) reconstruction within a shallow region under the illuminated surface of the turbid medium. The methodological foundation is laid on the single-pixel spatial frequency domain (SFD) imaging that facilitates the adoption of the well-established time-correlated single-photon counting (TCSPC)-based TD detection and generalized pulse spectrum techniques (GPST)-based reconstruction. To ameliorate the defects of the conventional diffusion equation (DE) in the forward modeling of TD-SFD-DOT, mainly the low accuracy in the near-field region and in profiling early-photon migration, we propose a modified model employing the time-dependent δ-P1 approximation and verify its improved accuracy in comparison with both the Monte Carlo and DE-based ones. For a simplified inversion process, a modified GPST approach is extended to TD-SFD-DOT that enables the effective separation of the absorption and scattering coefficients using a steady-state equivalent strategy. Furthermore, we set up a single-pixel TD-SFD-DOT system that employs the TCSPC-based TD detection in the SFD imaging framework. For assessments of the reconstruction approach and the system performance, phantom experiments are performed for a series of scenarios. The results show the effectiveness of the proposed methodology for rapid 3-D reconstruction of the absorption and scattering coefficients within a depth range of about 5 mean free pathlengths.

Application limits of the scaling relations for Monte Carlo simulations in diffuse optics. Part 1: theory

Monte Carlo (MC) is a powerful tool to study photon migration in scattering media, yet quite time-consuming to solve inverse problems. To speed up MC-simulations, scaling relations can be applied to an existing initial MC-simulation to generate a new data-set with different optical properties. We named this approach trajectory-based since it uses the knowledge of the detected photon trajectories of the initial MC-simulation, in opposition to the slower photon-based approach, where a novel MC-simulation is rerun with new optical properties. We investigated the convergence and applicability limits of the scaling relations, both related to the likelihood that the sample of trajectories considered is representative also for the new optical properties. For absorption, the scaling relation contains smoothly converging Lambert-Beer factors, whereas for scattering it is the product of two quickly diverging factors, whose ratio, for NIRS cases, can easily reach ten orders of magnitude. We investigated such instability by studying the probability-distribution for the number of scattering events in trajectories of given length. We propose a convergence test of the scattering scaling relation based on the minimum-maximum number of scattering events in recorded trajectories. We also studied the dependence of MC-simulations on optical properties, most critical in inverse problems, finding that scattering derivatives are ascribed to small deviations in the distribution of scattering events from a Poisson distribution. This paper, which can also serve as a tutorial, helps to understand the physics of the scaling relations with the causes of their limitations and devise new strategies to deal with them.

SpeckleCam: high-resolution computational speckle contrast tomography for deep blood flow imaging

Laser speckle contrast imaging is widely used in clinical studies to monitor blood flow distribution. Speckle contrast tomography, similar to diffuse optical tomography, extends speckle contrast imaging to provide deep tissue blood flow information. However, the current speckle contrast tomography techniques suffer from poor spatial resolution and involve both computation and memory intensive reconstruction algorithms. In this work, we present SpeckleCam, a camera-based system to reconstruct high resolution 3D blood flow distribution deep inside the skin. Our approach replaces the traditional forward model using diffuse approximations with Monte-Carlo simulations-based convolutional forward model, which enables us to develop an improved deep tissue blood flow reconstruction algorithm. We show that our proposed approach can recover complex structures up to 6 mm deep inside a tissue-like scattering medium in the reflection geometry. We also conduct human experiments to demonstrate that our approach can detect reduced flow in major blood vessels during vascular occlusion.

Effects of atlas-based anatomy on modelled light transport in the neonatal head

Objective.Diffuse optical tomography (DOT) provides a relatively convenient method for imaging haemodynamic changes related to neuronal activity on the cerebral cortex. Due to practical challenges in obtaining anatomical images of neonates, an anatomical framework is often created from an age-appropriate atlas model, which is individualized to the subject based on measurements of the head geometry. This work studies the approximation error arising from using an atlas instead of the neonate's own anatomical model.Approach.We consider numerical simulations of frequency-domain (FD) DOT using two approaches, Monte Carlo simulations and diffusion approximation via finite element method, and observe the variation in (1) the logarithm of amplitude and phase shift measurements, and (2) the corresponding inner head sensitivities (Jacobians), due to varying segmented anatomy. Varying segmentations are sampled by registering 165 atlas models from a neonatal database to the head geometry of one individual selected as the reference model. Prior to the registration, we refine the segmentation of the cerebrospinal fluid (CSF) by separating the CSF into two physiologically plausible layers.Main results.In absolute measurements, a considerable change in the grey matter or extracerebral tissue absorption coefficient was found detectable over the anatomical variation. In difference measurements, a small local 10%-increase in brain absorption was clearly detectable in the simulated measurements over the approximation error in the Jacobians, despite the wide range of brain maturation among the registered models.Significance.Individual-level atlas models could potentially be selected within several weeks in gestational age in DOT difference imaging, if an exactly age-appropriate atlas is not available. The approximation error method could potentially be implemented to improve the accuracy of atlas-based imaging. The presented CSF segmentation algorithm could be useful also in other model-based imaging modalities. The computation of FD Jacobians is now available in the widely-used Monte Carlo eXtreme software.

Analysis of relative error in perturbation Monte Carlo simulations of radiative transport

Significance

Perturbation and differential Monte Carlo (pMC/dMC) methods, used in conjunction with nonlinear optimization methods, have been successfully applied to solve inverse problems in diffuse optics. Application of pMC to systems over a large range of optical properties requires optimal "placement" of baseline conventional Monte Carlo (cMC) simulations to minimize the pMC variance. The inability to predict the growth in pMC solution uncertainty with perturbation size limits the application of pMC, especially for multispectral datasets where the variation of optical properties can be substantial.

Aim

We aim to predict the variation of pMC variance with perturbation size without explicit computation of perturbed photon weights. Our proposed method can be used to determine the range of optical properties over which pMC predictions provide sufficient accuracy. This method can be used to specify the optical properties for the reference cMC simulations that pMC utilizes to provide accurate predictions over a desired optical property range.

Approach

We utilize a conventional error propagation methodology to calculate changes in pMC relative error for Monte Carlo simulations. We demonstrate this methodology for spatially resolved diffuse reflectance measurements with ±20% scattering perturbations. We examine the performance of our method for reference simulations spanning a broad range of optical properties relevant for diffuse optical imaging of biological tissues. Our predictions are computed using the variance, covariance, and skewness of the photon weight, path length, and collision distributions generated by the reference simulation.

Results

We find that our methodology performs best when used in conjunction with reference cMC simulations that utilize Russian Roulette (RR) method. Specifically, we demonstrate that for a proximal detector placed immediately adjacent to the source, we can estimate the pMC relative error within 5% of the true value for scattering perturbations in the range of [ - 15 % , + 20 % ] . For a distal detector placed at ∼ 3 transport mean free paths relative to the source, our method provides relative error estimates within 20% for scattering perturbations in the range of [ - 8 % , + 15 % ] . Moreover, reference simulations performed at lower ( μ s ' / μ a ) values showed better performance for both proximal and distal detectors.

Conclusions

These findings indicate that reference simulations utilizing continuous absorption weighting (CAW) with the Russian Roulette method and executed using optical properties with a low ( μ s ' / μ a ) ratio spanning the desired range of μ s values, are highly advantageous for the deployment of pMC to obtain radiative transport estimates over a wide range of optical properties.

Monte Carlo Simulation of Diffuse Optical Spectroscopy for 3D Modeling of Dental Tissues

Three-dimensional precise models of teeth are critical for a variety of dental procedures, including orthodontics, prosthodontics, and implantology. While X-ray-based imaging devices are commonly used to obtain anatomical information about teeth, optical devices offer a promising alternative for acquiring 3D data of teeth without exposing patients to harmful radiation. Previous research has not examined the optical interactions with all dental tissue compartments nor provided a thorough analysis of detected signals at various boundary conditions for both transmittance and reflectance modes. To address this gap, a GPU-based Monte Carlo (MC) method has been utilized to assess the feasibility of diffuse optical spectroscopy (DOS) systems operating at 633 nm and 1310 nm wavelengths for simulating light-tissue interactions in a 3D tooth model. The results show that the system's sensitivity to detect pulp signals at both 633 nm and 1310 nm wavelengths is higher in the transmittance compared with that in the reflectance mode. Analyzing the recorded absorbance, reflectance, and transmittance data verified that surface reflection at boundaries can improve the detected signal, especially from the pulp region in both reflectance and transmittance DOS systems. These findings could ultimately lead to more accurate and effective dental diagnosis and treatment.

Deep learning-based fusion of widefield diffuse optical tomography and micro-CT structural priors for accurate 3D reconstructions

Widefield illumination and detection strategies leveraging structured light have enabled fast and robust probing of tissue properties over large surface areas and volumes. However, when applied to diffuse optical tomography (DOT) applications, they still require a time-consuming and expert-centric solving of an ill-posed inverse problem. Deep learning (DL) models have been recently proposed to facilitate this challenging step. Herein, we expand on a previously reported deep neural network (DNN) -based architecture (modified AUTOMAP - ModAM) for accurate and fast reconstructions of the absorption coefficient in 3D DOT based on a structured light illumination and detection scheme. Furthermore, we evaluate the improved performances when incorporating a micro-CT structural prior in the DNN-based workflow, named Z-AUTOMAP. This Z-AUTOMAP significantly improves the widefield imaging process's spatial resolution, especially in the transverse direction. The reported DL-based strategies are validated both in silico and in experimental phantom studies using spectral micro-CT priors. Overall, this is the first successful demonstration of micro-CT and DOT fusion using deep learning, greatly enhancing the prospect of rapid data-integration strategies, often demanded in challenging pre-clinical scenarios.

Unrolled-DOT: an interpretable deep network for diffuse optical tomography

SIGNIFICANCE

Imaging through scattering media is critical in many biomedical imaging applications, such as breast tumor detection and functional neuroimaging. Time-of-flight diffuse optical tomography (ToF-DOT) is one of the most promising methods for high-resolution imaging through scattering media. ToF-DOT and many traditional DOT methods require an image reconstruction algorithm. Unfortunately, this algorithm often requires long computational runtimes and may produce lower quality reconstructions in the presence of model mismatch or improper hyperparameter tuning.

AIM

We used a data-driven unrolled network as our ToF-DOT inverse solver. The unrolled network is faster than traditional inverse solvers and achieves higher reconstruction quality by accounting for model mismatch.

APPROACH

Our model "Unrolled-DOT" uses the learned iterative shrinkage thresholding algorithm. In addition, we incorporate a refinement U-Net and Visual Geometry Group (VGG) perceptual loss to further increase the reconstruction quality. We trained and tested our model on simulated and real-world data and benchmarked against physics-based and learning-based inverse solvers.

RESULTS

In experiments on real-world data, Unrolled-DOT outperformed learning-based algorithms and achieved over 10× reduction in runtime and mean-squared error, compared to traditional physics-based solvers.

CONCLUSION

We demonstrated a learning-based ToF-DOT inverse solver that achieves state-of-the-art performance in speed and reconstruction quality, which can aid in future applications for noninvasive biomedical imaging.

Sub-mm resolution tomographic imaging in turbid media by an ultra-high density multichannel approach

We demonstrate an ultra-high-density source-detector (SD) diffuse optical tomography system scalable to thousands of combinatorial SD pairs per cm3 of total voxel volume. We demonstrate the imaging of dynamic targets (including phantom arteries) with 100 um resolution at over 10 Hz frame rate within turbid media (> 60 MFP). Further, as a step toward a wearable mobile imager, we introduce monolithic mm-size dense semiconductor laser array chips as sources for potential unobtrusive epidermal tomographic use.

In Silico Investigation of SNR and Dermis Sensitivity for Optimum Dual-Channel Near-Infrared Glucose Sensor Designs for Different Skin Colors

Diabetes is a serious health condition that requires patients to regularly monitor their blood glucose level, making the development of practical, compact, and non-invasive techniques essential. Optical glucose sensors-and, specifically, NIR sensors-have the advantages of being non-invasive, compact, inexpensive, and user-friendly devices. However, these sensors have low accuracy and are yet to be adopted by healthcare providers. In our previous work, we introduced a non-invasive dual-channel technique for NIR sensors, in which a long channel is utilized to measure the glucose level in the inner skin (dermis) layer, while a short channel is used to measure the noise signal of the superficial skin (epidermis) layer. In this work, we investigated the use of dual-NIR channels for patients with different skin colors (i.e., having different melanin concentrations). We also adopted a Monte Carlo simulation model that takes into consideration the differences between different skin layers, in terms of blood content, water content, melanin concentration in the epidermis layer, and skin optical proprieties. On the basis of the signal-to-noise ratio, as well as the sensitivities of both the epidermis and dermis layers, we suggest the selection of wavelengths and source-to-detector separation for optimal NIR channels under different skin melanin concentrations. This work facilitates the improved design of a compact and non-invasive NIR glucose sensor that can be utilized by patients with different skin colors.

Signal and measurement considerations for human translation of diffuse in vivo flow cytometry

SIGNIFICANCE

"Diffuse in vivo flow cytometry" (DiFC) is an emerging technology for fluorescence detection of rare circulating cells directly in large deep-seated blood vessels in mice. Because DiFC uses highly scattered light, in principle, it could be translated to human use. However, an open question is whether fluorescent signals from single cells would be detectable in human-scale anatomies.

AIM

Suitable blood vessels in a human wrist or forearm are at a depth of ∼2 to 4 mm. The aim of this work was to study the impact of DiFC instrument geometry and wavelength on the detected DiFC signal and on the maximum depth of detection of a moving cell.

APPROACH

We used Monte Carlo simulations to compute fluorescence Jacobian (sensitivity) matrices for a range of source and detector separations (SDS) and tissue optical properties over the visible and near infrared spectrum. We performed experimental measurements with three available versions of DiFC (488, 640, and 780 nm), fluorescent microspheres, and tissue mimicking optical flow phantoms. We used both computational and experimental data to estimate the maximum depth of detection at each combination of settings.

RESULTS

For the DiFC detection problem, our analysis showed that for deep-seated blood vessels, the maximum sensitivity was obtained with NIR light (780 nm) and 3-mm SDS.

CONCLUSIONS

These results suggest that-in combination with a suitable molecularly targeted fluorescent probes-circulating cells and nanosensors could, in principle, be detectable in circulation in humans.

Adaptive stochastic Gauss-Newton method with optical Monte Carlo for quantitative photoacoustic tomography

SIGNIFICANCE

The image reconstruction problem in quantitative photoacoustic tomography (QPAT) is an ill-posed inverse problem. Monte Carlo method for light transport can be utilized in solving this image reconstruction problem.

AIM

The aim was to develop an adaptive image reconstruction method where the number of photon packets in Monte Carlo simulation is varied to achieve a sufficient accuracy with reduced computational burden.

APPROACH

The image reconstruction problem was formulated as a minimization problem. An adaptive stochastic Gauss-Newton (A-SGN) method combined with Monte Carlo method for light transport was developed. In the algorithm, the number of photon packets used on Gauss-Newton (GN) iteration was varied utilizing a so-called norm test.

RESULTS

The approach was evaluated with numerical simulations. With the proposed approach, the number of photon packets needed for solving the inverse problem was significantly smaller than in a conventional approach where the number of photon packets was fixed for each GN iteration.

CONCLUSIONS

The A-SGN method with a norm test can be utilized in QPAT to provide accurate and computationally efficient solutions.

Monte Carlo-based data generation for efficient deep learning reconstruction of macroscopic diffuse optical tomography and topography applications

SIGNIFICANCE

Deep learning (DL) models are being increasingly developed to map sensor data to the image domain directly. However, DL methodologies are data-driven and require large and diverse data sets to provide robust and accurate image formation performances. For research modalities such as 2D/3D diffuse optical imaging, the lack of large publicly available data sets and the wide variety of instrumentation designs, data types, and applications leads to unique challenges in obtaining well-controlled data sets for training and validation. Meanwhile, great efforts over the last four decades have focused on developing accurate and computationally efficient light propagation models that are flexible enough to simulate a wide variety of experimental conditions.

AIM

Recent developments in Monte Carlo (MC)-based modeling offer the unique advantage of simulating accurately light propagation spatially, temporally, and over an extensive range of optical parameters, including minimally to highly scattering tissue within a computationally efficient platform. Herein, we demonstrate how such MC platforms, namely "Monte Carlo eXtreme" and "Mesh-based Monte Carlo," can be leveraged to generate large and representative data sets for training the DL model efficiently.

APPROACH

We propose data generator pipeline strategies using these platforms and demonstrate their potential in fluorescence optical topography, fluorescence optical tomography, and single-pixel diffuse optical tomography. These applications represent a large variety in instrumentation design, sample properties, and contrast function.

RESULTS

DL models trained using the MC-based in silico datasets, validated further with experimental data not used during training, show accurate and promising results.

CONCLUSION

Overall, these MC-based data generation pipelines are expected to support the development of DL models for rapid, robust, and user-friendly image formation in a wide variety of applications.

A Monte Carlo study of near infrared light propagation in the human head with lesions-a time-resolved approach

Several clinical conditions leading to traumatic brain injury can cause hematomas or edemas inside the cerebral tissue. If these are not properly treated in time, they are prone to produce long-term neurological disabilities, or even death. Low-cost, portable and easy-to-handle devices are desired for continuous monitoring of these conditions and Near Infrared Spectroscopy (NIRS) techniques represent an appropriate choice. In this work, we use Time-Resolved (TR) Monte Carlo simulations to present a study of NIR light propagation over a digital MRI phantom. Healthy and injured (hematoma/edema) situations are considered. TR Diffuse Reflectance simulations for different lesion volumes and interoptode distances are performed in the frontal area and the left parietal area. Results show that mean partial pathlengths, photon measurement density functions and time dependent contrasts are sensitive to the presence of lesions, allowing their detection mainly for intermediate optodes separations, which proves that these metrics represent robust means of diagnose and monitoring. Conventional Continuous Wave (CW) contrasts are also presented as a particular case of the time dependent ones, but they result less sensitive to the lesions, and have higher associated uncertainties.

High-resolution three-dimensional blood flow tomography in the subdiffuse regime using laser speckle contrast imaging

SIGNIFICANCE

Visualizing high-resolution hemodynamics in cerebral tissue over a large field of view (FOV), provides important information in studying disease states affecting the brain. Current state-of-the-art optical blood flow imaging techniques either lack spatial resolution or are too slow to provide high temporal resolution reconstruction of flow map over a large FOV.

AIM

We present a high spatial resolution computational optical imaging technique based on principles of laser speckle contrast imaging (LSCI) for reconstructing the blood flow maps in complex tissue over a large FOV provided that the three-dimensional (3D) vascular structure is known or assumed.

APPROACH

Our proposed method uses a perturbation Monte Carlo simulation of the high-resolution 3D geometry for both accurately deriving the speckle contrast forward model and calculating the Jacobian matrix used in our reconstruction algorithm to achieve high resolution. Given the convex nature of our highly nonlinear problem, we implemented a mini-batch gradient descent with an adaptive learning rate optimization method to iteratively reconstruct the blood flow map. Specifically, we implemented advanced optimization techniques combined with efficient parallelization and vectorization of the forward and derivative calculations to make reconstruction of the blood flow map feasible with reconstruction times on the order of tens of minutes.

RESULTS

We tested our reconstruction algorithm through simulation of both a flow phantom model as well as an anatomically correct murine cerebral tissue and vasculature captured via two-photon microscopy. Additionally, we performed a noise study, examining the robustness of our inverse model in presence of 0.1% and 1% additive noise. In all cases, the blood flow reconstruction error was <2  %   for most of the vasculature, except for the peripheral vasculature which suffered from insufficient photon sampling. Descending vasculature and deeper structures showed slightly higher sensitivity to noise compared with vasculature with a horizontal orientation at the more superficial layers. Our results show high-resolution reconstruction of the blood flow map in tissue down to 500  μm and beyond.

CONCLUSIONS

We have demonstrated a high-resolution computational imaging technique for visualizing blood flow map in complex tissue over a large FOV. Once a high-resolution structural image is captured, our reconstruction algorithm only requires a few LSCI images captured through a camera to reconstruct the blood flow map computationally at a high resolution. We note that the combination of high temporal and spatial resolution of our reconstruction algorithm makes the solution well-suited for applications involving fast monitoring of flow dynamics over a large FOV, such as in functional neural imaging.

Imaging affective and non-affective touch processing in two-year-old children

Touch is an important component of early parent-child interaction and plays a critical role in the socio-emotional development of children. However, there are limited studies on touch processing amongst children in the age range from one to three years. The present study used frequency-domain diffuse optical tomography (DOT) to investigate the processing of affective and non-affective touch over left frontotemporal brain areas contralateral to the stimulated forearm in two-year-old children. Affective touch was administered by a single stroke with a soft brush over the child's right dorsal forearm at 3 cm/s, while non-affective touch was provided by multiple brush strokes at 30 cm/s. We found that in the insula, the total haemoglobin (HbT) response to slow brushing was significantly greater than the response to fast brushing (slow > fast). Additionally, a region in the postcentral gyrus, Rolandic operculum and superior temporal gyrus exhibited greater response to fast brushing than slow brushing (fast > slow). These findings confirm that an adult-like pattern of haemodynamic responses to affective and non-affective touch can be recorded in two-year-old subjects using DOT. To improve the accuracy of modelling light transport in the two-year-old subjects, we used a published age-appropriate atlas and deformed it to match the exterior shape of each subject's head. We estimated the combined scalp and skull, and grey matter (GM) optical properties by fitting simulated data to calibrated and coupling error corrected phase and amplitude measurements. By utilizing a two-compartment cerebrospinal fluid (CSF) model, the accuracy of estimation of GM optical properties and the localization of activation in the insula was improved. The techniques presented in this paper can be used to study neural development of children at different ages and illustrate that the technology is well-tolerated by most two-year-old children and not excessively sensitive to subject movement. The study points the way towards exciting possibilities in functional imaging of deeper functional areas near sulci in small children.

MOCA: a systematic toolbox for designing and assessing modular functional near-infrared brain imaging probes

SIGNIFICANCE

The expansion of functional near-infrared spectroscopy (fNIRS) systems toward broader utilities has led to the emergence of modular fNIRS systems composed of repeating optical source/detector modules. Compared to conventional fNIRS systems, modular fNIRS systems are more compact and flexible, making wearable and long-term monitoring possible. However, the large number of design parameters makes understanding their impact on a probe's performance a daunting task.

AIM

We aim to create a systematic software platform to facilitate the design, characterization, and comparison of modular fNIRS probes.

APPROACH

Our software-modular optode configuration analyzer (MOCA)-implements semi-automatic algorithms that assist in tessellating user-specified regions-of-interest, in interconnecting modules of various shapes, and in quantitatively comparing probe performance using metrics, such as spatial channel distributions and average brain sensitivity of the resulting probes. There is also support for limited parameter sweeping capabilities.

RESULTS

Through several examples, we show that users can use MOCA to design and optimize modular fNIRS probes, study trade-offs between several module shapes, improve brain sensitivity in probes via module re-orientation, and enhance probe performance via adjusting module spatial layouts.

CONCLUSION

Despite its simplicity, our modular probe design platform offers a framework to describe and quantitatively assess probes made by modules, opening a new door for the growing fNIRS user community to approach the challenging problem of module- and probe-parameter selection and fine-tuning.

MCX Cloud-a modern, scalable, high-performance and in-browser Monte Carlo simulation platform with cloud computing

SIGNIFICANCE

Despite the ample progress made toward faster and more accurate Monte Carlo (MC) simulation tools over the past decade, the limited usability and accessibility of these advanced modeling tools remain key barriers to widespread use among the broad user community.

AIM

An open-source, high-performance, web-based MC simulator that builds upon modern cloud computing architectures is highly desirable to deliver state-of-the-art MC simulations and hardware acceleration to general users without the need for special hardware installation and optimization.

APPROACH

We have developed a configuration-free, in-browser 3D MC simulation platform-Monte Carlo eXtreme (MCX) Cloud-built upon an array of robust and modern technologies, including a Docker Swarm-based cloud-computing backend and a web-based graphical user interface (GUI) that supports in-browser 3D visualization, asynchronous data communication, and automatic data validation via JavaScript Object Notation (JSON) schemas.

RESULTS

The front-end of the MCX Cloud platform offers an intuitive simulation design, fast 3D data rendering, and convenient simulation sharing. The Docker Swarm container orchestration backend is highly scalable and can support high-demand GPU MC simulations using MCX over a dynamically expandable virtual cluster.

CONCLUSION

MCX Cloud makes fast, scalable, and feature-rich MC simulations readily available to all biophotonics researchers without overhead. It is fully open-source and can be freely accessed at http://mcx.space/cloud.

A Data Self-Calibration Method Based on High-Density Parallel Plate Diffuse Optical Tomography for Breast Cancer Imaging

When performing the diffuse optical tomography (DOT) of the breast, the mismatch between the forward model and the experimental conditions will significantly hinder the reconstruction accuracy. Therefore, the reference measurement is commonly used to calibrate the measured data before the reconstruction. However, it is complicated to customize corresponding reference phantoms based on the breast shape and background optical parameters of different subjects in clinical trials. Furthermore, although high-density (HD) DOT configuration has been proven to improve imaging quality, a large number of source-detector (SD) pairs also increase the difficulty of multi-channel correction. To enhance the applicability of the breast DOT, a data self-calibration method based on an HD parallel-plate DOT system is proposed in this paper to replace the conventional relative measurement on a reference phantom. The reference predicted data can be constructed directly from the measurement data with the support of the HD-DOT system, which has nearly a hundred sets of measurements at each SD distance. The proposed scheme has been validated by Monte Carlo (MC) simulation, breast-size phantom experiments, and clinical trials, exhibiting the feasibility in ensuring the quality of the DOT reconstruction while effectively reducing the complexity associated with relative measurements on reference phantoms.

Deep Learning in Biomedical Optics

This article reviews deep learning applications in biomedical optics with a particular emphasis on image formation. The review is organized by imaging domains within biomedical optics and includes microscopy, fluorescence lifetime imaging, in vivo microscopy, widefield endoscopy, optical coherence tomography, photoacoustic imaging, diffuse tomography, and functional optical brain imaging. For each of these domains, we summarize how deep learning has been applied and highlight methods by which deep learning can enable new capabilities for optics in medicine. Challenges and opportunities to improve translation and adoption of deep learning in biomedical optics are also summarized. Lasers Surg. Med. © 2021 Wiley Periodicals LLC.

High Resolution, Deep Imaging Using Confocal Time-of-Flight Diffuse Optical Tomography

Light scattering by tissue severely limits how deep beneath the surface one can image, and the spatial resolution one can obtain from these images. Diffuse optical tomography (DOT) is one of the most powerful techniques for imaging deep within tissue - well beyond the conventional  ∼ 10-15 mean scattering lengths tolerated by ballistic imaging techniques such as confocal and two-photon microscopy. Unfortunately, existing DOT systems are limited, achieving only centimeter-scale resolution. Furthermore, they suffer from slow acquisition times and slow reconstruction speeds making real-time imaging infeasible. We show that time-of-flight diffuse optical tomography (ToF-DOT) and its confocal variant (CToF-DOT), by exploiting the photon travel time information, allow us to achieve millimeter spatial resolution in the highly scattered diffusion regime ( mean free paths). In addition, we demonstrate two additional innovations: focusing on confocal measurements, and multiplexing the illumination sources allow us to significantly reduce the measurement acquisition time. Finally, we rely on a novel convolutional approximation that allows us to develop a fast reconstruction algorithm, achieving a 100× speedup in reconstruction time compared to traditional DOT reconstruction techniques. Together, we believe that these technical advances serve as the first step towards real-time, millimeter resolution, deep tissue imaging using DOT.

Light transport modeling in highly complex tissues using the implicit mesh-based Monte Carlo algorithm

The mesh-based Monte Carlo (MMC) technique has grown tremendously since its initial publication nearly a decade ago. It is now recognized as one of the most accurate Monte Carlo (MC) methods, providing accurate reference solutions for the development of novel biophotonics techniques. In this work, we aim to further advance MMC to address a major challenge in biophotonics modeling, i.e. light transport within highly complex tissues, such as dense microvascular networks, porous media and multi-scale tissue structures. Although the current MMC framework is capable of simulating light propagation in such media given its generality, the run-time and memory usage grow rapidly with increasing media complexity and size. This greatly limits our capability to explore complex and multi-scale tissue structures. Here, we propose a highly efficient implicit mesh-based Monte Carlo (iMMC) method that incorporates both mesh- and shape-based tissue representations to create highly complex yet memory-efficient light transport simulations. We demonstrate that iMMC is capable of providing accurate solutions for dense vessel networks and porous tissues while reducing memory usage by greater than a hundred- or even thousand-fold. In a sample network of microvasculature, the reduced shape complexity results in nearly 3x speed acceleration. The proposed algorithm is now available in our open-source MMC software at http://mcx.space/#mmc.

Perturbation Monte Carlo Method for Quantitative Photoacoustic Tomography

Quantitative photoacoustic tomography aims at estimating optical parameters from photoacoustic images that are formed utilizing the photoacoustic effect caused by the absorption of an externally introduced light pulse. This optical parameter estimation is an ill-posed inverse problem, and thus it is sensitive to measurement and modeling errors. In this work, we propose a novel way to solve the inverse problem of quantitative photoacoustic tomography based on the perturbation Monte Carlo method. Monte Carlo method for light propagation is a stochastic approach for simulating photon trajectories in a medium with scattering particles. It is widely accepted as an accurate method to simulate light propagation in tissues. Furthermore, it is numerically robust and easy to implement. Perturbation Monte Carlo maintains this robustness and enables forming gradients for the solution of the inverse problem. We validate the method and apply it in the framework of Bayesian inverse problems. The simulations show that the perturbation Monte Carlo method can be used to estimate spatial distributions of both absorption and scattering parameters simultaneously. These estimates are qualitatively good and quantitatively accurate also in parameter scales that are realistic for biological tissues.

Depth-resolved imaging of photosensitizer in the rodent brain using fluorescence laminar optical tomography

SIGNIFICANCE

Previous studies have been performed to image photosensitizers in certain organs and tumors using fluorescence laminar optical tomography. Currently, no work has yet been published to quantitatively compare the signal compensation of fluorescence laminar optical tomography with two-dimensional (2-D) imaging in tissues.

AIM

The purpose of this study is to quantify the benefit that fluorescence laminar optical tomography holds over 2-D imaging. We compared fluorescence laminar optical tomography with maximum intensity projection imaging to simulate 2-D imaging, as this would be the most similar and stringent comparison.

APPROACH

A capillary filled with a photosensitizer was placed in a phantom and ex vivo rodent brains, with fluorescence laminar optical tomography and maximum intensity projection images obtained. The signal loss in the Z direction was quantified and compared to see which methodology could compensate better for signal loss caused by tissue attenuation.

RESULTS

The results demonstrated that we can reconstruct a capillary filled with benzoporphyrin derivative photosensitizers faithfully in phantoms and in ex vivo rodent brain tissues using fluorescence laminar optical tomography. We further demonstrated that we can better compensate for signal loss when compared with maximum intensity projection imaging.

CONCLUSIONS

Using fluorescence laminar optical tomography (FLOT), one can compensate for signal loss in deeper parts of tissue when imaging in ex vivo rodent brain tissue compared with maximum intensity projection imaging.

Macroscopic fluorescence lifetime topography enhanced via spatial frequency domain imaging

We report on a macroscopic fluorescence lifetime imaging (MFLI) topography computational framework based around machine learning with the main goal of retrieving the depth of fluorescent inclusions deeply seated in bio-tissues. This approach leverages the depth-resolved information inherent to time-resolved fluorescence data sets coupled with the retrieval of in situ optical properties as obtained via spatial frequency domain imaging (SFDI). Specifically, a Siamese network architecture is proposed with optical properties (OPs) and time-resolved fluorescence decays as input followed by simultaneous retrieval of lifetime maps and depth profiles. We validate our approach using comprehensive in silico data sets as well as with a phantom experiment. Overall, our results demonstrate that our approach can retrieve the depth of fluorescence inclusions, especially when coupled with optical properties estimation, with high accuracy. We expect the presented computational approach to find great utility in applications such as optical-guided surgery.

Efficient inversion strategies for estimating optical properties with Monte Carlo radiative transport models

SIGNIFICANCE

Indirect imaging problems in biomedical optics generally require repeated evaluation of forward models of radiative transport, for which Monte Carlo is accurate yet computationally costly. We develop an approach to reduce this bottleneck, which has significant implications for quantitative tomographic imaging in a variety of medical and industrial applications.

AIM

Our aim is to enable computationally efficient image reconstruction in (hybrid) diffuse optical modalities using stochastic forward models.

APPROACH

Using Monte Carlo, we compute a fully stochastic gradient of an objective function for a given imaging problem. Leveraging techniques from the machine learning community, we then adaptively control the accuracy of this gradient throughout the iterative inversion scheme to substantially reduce computational resources at each step.

RESULTS

For example problems of quantitative photoacoustic tomography and ultrasound-modulated optical tomography, we demonstrate that solutions are attainable using a total computational expense that is comparable to (or less than) that which is required for a single high-accuracy forward run of the same Monte Carlo model.

CONCLUSIONS

This approach demonstrates significant computational savings when approaching the full nonlinear inverse problem of optical property estimation using stochastic methods.

Accelerating Monte Carlo modeling of structured-light-based diffuse optical imaging via "photon sharing"

The increasing use of spatially modulated imaging and single-pixel detection techniques demands computationally efficient methods for light transport modeling. Herein, we report an easy-to-implement yet significantly more efficient Monte Carlo (MC) method for simultaneously simulating spatially modulated illumination and detection patterns accurately in 3D complex domains. We have implemented this accelerated algorithm, named "photon sharing," in our open-source MC simulators, reporting 13.6× and 5.5× speedups in mesh- and voxel-based MC benchmarks, respectively. In addition, the proposed algorithm is readily used to accelerate the solving of inverse problems in spatially modulated imaging systems by building Jacobians of all illumination-detection pattern pairs concurrently, resulting in a 12.4-fold speed improvement.

System configuration optimization for mesoscopic fluorescence molecular tomography

Tissue engineering applications demand 3D, non-invasive, and longitudinal assessment of bioprinted constructs. Current emphasis is on developing tissue constructs mimicking in vivo conditions; however, these are increasingly challenging to image as they are typically a few millimeters thick and turbid, limiting the usefulness of classical fluorescence microscopic techniques. For such applications, we developed a Mesoscopic Fluorescence Molecular Tomography methodology that collects high information content data to enable high-resolution tomographic reconstruction of fluorescence biomarkers at millimeters depths. This imaging approach is based on an inverse problem; hence, its imaging performances are dependent on critical technical considerations including optode sampling, forward model design and inverse solver parameters. Herein, we investigate the impact of the optical system configuration parameters, including detector layout, number of detectors, combination of detector and source numbers, and scanning mode with uncoupled or coupled source and detector array, on the 3D imaging performances. Our results establish that an MFMT system with a 2D detection chain implemented in a de-scanned mode provides the optimal imaging reconstruction performances.

Graphics processing unit-accelerated mesh-based Monte Carlo photon transport simulations

The mesh-based Monte Carlo (MMC) algorithm is increasingly used as the gold-standard for developing new biophotonics modeling techniques in 3-D complex tissues, including both diffusion-based and various Monte Carlo (MC)-based methods. Compared to multilayered and voxel-based MCs, MMC can utilize tetrahedral meshes to gain improved anatomical accuracy but also results in higher computational and memory demands. Previous attempts of accelerating MMC using graphics processing units (GPUs) have yielded limited performance improvement and are not publicly available. We report a highly efficient MMC-MMCL-using the OpenCL heterogeneous computing framework and demonstrate a speedup ratio up to 420× compared to state-of-the-art single-threaded CPU simulations. The MMCL simulator supports almost all advanced features found in our widely disseminated MMC software, such as support for a dozen of complex source forms, wide-field detectors, boundary reflection, photon replay, and storing a rich set of detected photon information. Furthermore, this tool supports a wide range of GPUs/CPUs across vendors and is freely available with full source codes and benchmark suites at http://mcx.space/#mmc.

Head model based on the shape of the subject's head for optical brain imaging

Optical imaging methods such as near-infrared spectroscopy and diffuse optical tomography rely on models to solve the inverse problem. Imaging an adult human head also requires a head model. Using a model, which makes describing the structure of the head better, leads to acquiring a more accurate absorption map. Here, by combining the key features of layered slab models and head atlases, we introduce a new two-layered head model that is based on the surface geometry of the subject's head with variable thickness of the superficial layer. Using the Monte Carlo approach, we assess the performance of our model for fitting the optical properties from simulated time-resolved data of the adult head in a null distance source-detector configuration. Using our model, we observed improved results at 70 percent of the locations on the head and an overall 20 percent reduction in relative error compared to layered slab model.

Time-Domain Near-Infrared Spectroscopy and Imaging: A Review

This article reviews the past and current statuses of time-domain near-infrared spectroscopy (TD-NIRS) and imaging. Although time-domain technology is not yet widely employed due to its drawbacks of being cumbersome, bulky, and very expensive compared to commercial continuous wave (CW) and frequency-domain (FD) fNIRS systems, TD-NIRS has great advantages over CW and FD systems because time-resolved data measured by TD systems contain the richest information about optical properties inside measured objects. This article focuses on reviewing the theoretical background, advanced theories and methods, instruments, and studies on clinical applications for TD-NIRS including some clinical studies which used TD-NIRS systems. Major events in the development of TD-NIRS and imaging are identified and summarized in chronological tables and figures. Finally, prospects for TD-NIRS in the near future are briefly described.