This PDF file contains the front matter associated with SPIE Proceedings Volume 11074, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

We present here a novel time-domain diffuse optical detection chain consisting of a large area Silicon PhotoMultipliers (SiPM) coupled to a high count-rate timing electronics (TimeHarp 260 PICO) to achieve sustainable count-rates up to 10 Mcps without significant distortions to the distribution of time-of-flight (DTOF). Thanks to the large area of the detector (9 mm²) and the possibility to directly place it in contact with the sample (thus achieving a numerical aperture close to unity), the photon collection efficiency of the proposed detection chain is almost two orders of magnitude higher than traditional fiber-mounted PMT-based systems. This allows the detection also of the few late photons coming from deeper layers at short acquisition times, thus improving the robustness of the detection of localized inhomogeneities. We then demonstrate that, despite the high dark count rate of the detector, it is possible to reliably extract the optical properties of calibrated phantoms, with proper linearity and accuracy. We also explore the capability of the new detection chain for detecting brain activations. This work opens up the possibility of ultimate performance in terms of high signal and photon throughput, with compact, low cost, relatively simple front-end electronics detector coupled to innovative timing electronics, with exciting opportunities to expand it to tomographic applications.

We discuss advances in and applications of fibre-less, wearable, high-density diffuse optical tomography technologies, including a new device specifically for the newborn infant that employs flex-rigid PCB technology and provides channel density approaching 10 channels/cm².

We present a wearable TD-NIRS system (two wavelengths, one channel). The system is battery operated, can be remotely controlled and is able to perform measurements on brain and muscle on freely-moving subjects.

The next generation of diffuse optical imaging systems will consist of wearable and fiber-less devices, to exploit the advantages of diffuse optical imaging over other functional neuroimaging techniques and meet the needs of users to acquire data in real-world settings. Recently, research at UCL gave rise to a novel, modular high-density diffuse optical tomography (DOT) system that was validated by reconstructing activation images over the motor cortex of a thumb-tofinger extension task. The real question, however, is whether these fiber-less systems can be employed whilst the subject performs real-world activities, that is, whether they can provide reliable signals during participant motion. Integrating motion sensors into modular wearable electronics is straightforward. In this study we acquired DOT and motion sensor data whilst participants performed different activities involving motion. In one acquisition, only accelerometer data were acquired while in the second acquisition, all 9-axis of data (accelerometer, gyroscope and magnetometer data) were acquired. Results demonstrated that acceleration data from motion sensors is not enough to detect motion artifacts whilst performing active movement (e.g., walking), since the global motion obscures any subtle motion artifact. Conversely, by combining accelerometer and gyroscope data it seems possible to detect motion artifacts even during walking, that is when a global motion is present. However, not all types of motion artifacts (e.g., eyebrow raising) could be detected even with this full data configuration. Further studies are required to shed light on this important research question.

We present a newly developed multichannel broadband NIRS (or bNIRS) system that has the capacity to measure changes in light attenuation of 308 NIR wavelengths (610nm to 918nm) simultaneously over 16 different brain locations. To achieve this the instrument uses a lens based spectrometer with a front-illuminated CCD that has a sensor size of 26.8x26mm. This large CCD detector allows the simultaneous binning of 16 detector fibres. The software uses the UCLn algorithm to quantify the changes in oxy-, deoxy- haemoglobin concentration (HbO₂, HHb) and oxidised cytochrome-coxidase (oxCCO) simultaneously over 16 different brain locations with 1second sampling rate. We demonstrate the use of the instrument in quantifying brain tissue oxygenation and metabolic activity simultaneously with electrical changes as measured with EEG in children with seizures.

Diffuse optical tomography (DOT) images the distribution of the optical properties, such as the absorption and scattering coefficients, via the image reconstruction from the light intensities measured at the surface of the biological medium. The changes in the optical properties reflect the conditions of the tissues. Therefore, DOT image can provide the information which is not obtained from the other modalities and is useful for medical diagnoses. In this study, the application of the DOT to thyroid cancer diagnosis was investigated. The ultrasound technique is usually carried out for the thyroid cancer diagnosis. It is, however, difficult to distinguish follicular carcinoma from adenoma of thyroid. The optical properties may be helpful for the diagnosis. The image reconstruction algorithm employing the regularization minimizing l_p-norm (0 < p < 2) of the reconstructed image was developed. The image was reconstructed from the timeresolved measurement data. The numerical simulations of the image reconstruction were tried. The numerical simulation demonstrated that the developed algorithm was able to image the changes in the optical properties in the medium. Additionally, the image reconstruction of the numerical neck phantom was simulated. The thyroid cancer region was reconstructed successfully. It was demonstrated that the developed algorithm had the possibility to image thyroid cancer.

Time Domain Diffuse Optical Tomography (TD-DOT) performed at multiple wavelengths can be used to non-invasively probe tissue composition. Then, tissue composition can be related to breast tissue and lesion type. Thus, TD-DOT could be used for therapy monitoring for breast cancer. We developed a software tool for multi-wavelength TD-DOT and performed a validation on meat phantoms to mimic tissue heterogeneity. An inclusion of different meat was exploited to mimic the presence of a lesion in the breast. Results show good localization of the inclusion, but poor quantification of the reconstructed breast composition. The use of a morphological prior constraint, providing information on inclusion geometry and position, significantly improves both localization and composition estimate.

Diffuse optical tomography (DOT) estimates the optical properties inside a turbid medium by injecting light from the surface and measuring the reflected photons. In time-resolved technology, since to perform DOT reconstruction at time domain is too computationally expensive, datatypes are used instead. Temporal windows are the most used datatypes but until now just w(t) = tⁿe−^pt forms could be computed fast. In this work, we propose a new method to compute efficiently a larger set of window datatypes. The results show that with these new windows (1) the localization of inclusions deeper than 2.5 cm is improved and (2) the absorption quantification is ameliorated at all inclusion depths.

A multivariate method integrating time-domain and space-domain techniques of near-infrared spectroscopy is proposed for simultaneously retrieving the absolute quantities of optical scattering and absorption properties in tissues. The concept is theoretically demonstrated by Monte-Carlo simulations in the homogenous case, and then applied on twolayer liquid phantoms. The deviations from nominal values are typically less than 6% for absorption coefficients in both layers

We apply time-domain diffuse correlation spectroscopy (TD DCS) to quantify dynamics in a two-layer turbid phantom, where the top layer contains purely static, while bottom layer contains only dynamic scatterers. We demonstrate that the standard TD DCS processing is incapable to properly quantify dynamics at short sourcedetector distances (<1 cm) due to strong influence of the static layer. To solve this problem we introduce a novel model accompanied by the numerical noise-correction, which allows to properly recover the autocorrelation decay of the dynamic homogeneous medium hidden by the static turbid layer. Our approach can be thus beneficial for DCS applications in samples with mixed dynamics.

Fluorescence Molecular Tomography (FMT) constitutes a functional and molecular imaging tool complementing well-established anatomical imaging modalities such as magnetic resonance imaging (MRI) and X-ray computer tomography (CT). Preclinical applications of FMT in the fields of tumor imaging, breast imaging, proteomic research, and drug discovery have been reported. We have developed the Smart ToolkIt for Fluorescence Tomography (STIFT) software platform for simulating, reconstructing and optimizing FMT. However, accurate reconstruction of FMT data for inhomogeneous objects of complex shape, e.g. a live mouse, requires prior information about the anatomy. This work addresses this issue by incorporating a digital mouse atlas into STIFT. The Digimouse consists of high-resolution and well-registered multimodal volumetric images of a post-mortem 28g male C57Bl/6 mouse. Techniques including cropping, translation, and de-noising were performed on the Digimouse atlas (DA), the Digimouse CT volume (DCT), and our MR datasets. To decrease the discrepancy between the DCT and the DA, the skull regions of these two volumes were enhanced and the DA was deformed using the transformation matrix. Followed, was another automated registration, between an experimental MRI dataset and the newly coupled DCT-DA dataset by fixing the former. Finally, the registered DA was used for meshing and FMT simulations. The quality of registration was quantified and shown to have increased organ volume overlap using the Dice similarity coefficient as metric. The reconstructed fluorophore distribution matches well with the preset ground truth, demonstrating that the meshed atlas is compatible with STIFT.

Critical glycemic events, such as hypo- or hyperglycemia, are extremely common during the first week post-partum in very preterm neonates. Both hypo- and hyperglycemic changes have been associated with poor neurological outcome. Continuous glucose monitoring (CGM) is a promising tool to reduce glycemic variability in the preterm population and whole-head Diffuse Optical Tomography (DOT) is a promising tool for continuous monitoring of brain hemodynamics in newborns. In this study, we performed a combined CGM-DOT acquisition in a very preterm newborn (28 weeks gestational age). The newborn was monitored for 7 days continuously. Twelve events were detected during this period: 8 mild hypoglycemic events, one severe hypoglycemic event, two mild hyperglycemic events and one event with a mild hypo- followed by a mild-hyperglycemia. DOT data were available for all the events but two. DOT data were reconstructed with a neonatal head model for the severe hypoglycemic event before the start of the hypoglycemic event and during the maximum peak of hypoglycemia. These preliminary results showed a regional specificity of the hemodynamic changes during hypoglycemia, with a predominant recruitment of the motor and parietal areas. This study highlights the importance of using whole-head DOT in this research field and the feasibility to perform combined CGMDOT monitoring in very preterm neonates. Future clinical trials are required to investigate this clinical problem more thoroughly and shed light on the impact of tight glycemic control on the newborn brain.

We present pilot results on the validation of non-invasive assessment of elevated intracranial pressure with optical measurement of critical closing pressure. A strong correlation (r=0.85) between optical measurements of critical closing pressure and invasive measurements of intracranial pressure was observed in 5 infants with hydrocephalus, and 1 adult patient with diffuse hypoxic ischemic brain injury. By facilitating timely detection of intracranial hypertension, this approach has potential to reduce risk of brain damage in hydrocephalus and other vulnerable patient populations.

Using naturalistic language generation tasks (e.g. overt speech) to capture the neural correlates of speech production is important, as less naturalistic (but often used) tasks such as covert language generation are not a reliable substitute for accurately assessing cortical activation associated with naturalistic speech. fMRI poses challenges to implementing naturalistic language tasks, especially in clinical populations, because it is noisy, physically constraining, and contraindicated in populations with metal implants. High-density diffuse optical tomography (HD-DOT) is particularly well-suited for naturalistic language tasks because it is silent, wearable, portable, and metal-compatible. This study investigates cortical activity underlying naturalistic language generation using HD-DOT. Six adult subjects aged 20-26 years completed two scans on two separate days consisting of three different tasks: covert word reading (RW), covert verb generation (CV), and overt verb generation (OV). Cortical responses were apparent in expected anatomical areas for all tasks and RW, CV, and OV evoked responses of increasing strength (peak ΔHbO (μMol) = 7.58, 10.3, and 11.0, respectively). Notably, OV recruits additional activation in Broca’s area and right-lateralized primary motor cortex as compared to CV. These findings are consistent with those obtained using fMRI^9,10,11 and underscore the need to use naturalistic language tasks when assessing the neural representations of natural speech. These results motivate extension to further investigations of naturalistic language processing of increased complexity, both receptive and productive, such as within-room conversation.

A cloud-based NIRFAST server has been developed as a standalone tool for recovery of human thyroid hemodynamic parameters. The methodology utilize data from three complimentary diagnostic techniques: multi-spectral time-resolved spectroscopy (TRS) to measure concentrations of tissue constituents, diffuse correlation spectroscopy (DCS) to assess perfusion and ultrasound (US) images for optical probes guidance and building the tissue model. The recovery procedures benefit from using finite elements modelling and data processing based on highly massive parallel algorithms running on graphical processors with NVidia CUDA technology. The total execution time of thyroid hemodynamic parameters recovery is <120s. The server communication and calculation procedures are user-configurable through setting files. Therefore, the modular software architecture and remote access through a network interface make the server a universal tool applicable within a range of biomedical optics applications. The server can work as a post-processing box within an instrumental environment as shown in this work, a WWW-based software as a service or as a laboratory server. The NIRFAST server and methodology of analysing TRS/DCS/US data tuple were tested in-vivo on two thyroid malignant cancer patients showing differentiation in hemodynamic parameters (oxygenation and perfusion) between malignant and healthy tissue.

Multiple challenges exist in standardization of data format and processing pipelines for optical functional neuroimaging. We present here a fully self-contained and end-user-friendly tool that provides the flexibility for multiple array-based imaging modalities, and offers format compatibility, spatial registration, and analytical breadth and sophistication for postprocessing. NeuroDOT, available on GitHub, is written in MATLAB in the style of a conventional MATLAB toolbox with functionality distributed across several pipelines with extensive functions for data quality analysis and visualization. To aid in end-user support at multiple levels of familiarity and expertise, beyond the basic functionality, NeuroDOT contains data samples, support files, help sections, appendices, and tutorials. Anonymized and published data samples have been chosen to reflect common experimental paradigms in neuroimaging (e.g., retinotopy and language based tasks), and are provided in both raw and pre-processed versions to aid in troubleshooting and training for the new user. The NeuroDOT toolbox currently supports a wide variety of standard data file formats (e.g., NIFTI, GIFTI, and others). Help sections exist for each function and are searchable from the MATLAB command line, with a Help Viewer version as well. Both are written and formatted in the style of their native MATLAB counterparts for familiarity and ease of use. Several appendices detail data structures, pipelines and their construction, and select visualizations of our pipelines’ results for multiple data samples. Several tutorials are also included, each of which runs a data sample through a given pipeline to help the user harness the power and flexibility of NeuroDOT.

A software for fast rendering the visual appearance of a blood vessel in human skin is presented. Possibilities and Limitations of solving the inverse problem to obtain physiological information are discussed.

Brain tissue oxygen saturation, StO₂, measured with near-infrared spectroscopy (NIRS) is of great clinical interest as it quantifies the balance between cerebral oxygen supply and demand. Some brain oximeters are based on spatially resolved spectroscopy (SRS), where NIRS data is collected at multiple distances from the light source to estimate a slope of light attenuation against distance. Other use a broadband approach which utilizes derivatives of the absorption spectra to estimate StO₂, such as broadband fitting (BF). We describe a novel algorithm, broadband spatially resolved spectroscopy (BB-SRS), for estimating StO₂. It is based on comparing the measured slope to a model of the attenuation slope, which depends on the optical properties of tissue. Fitting this model with a least squares fitting procedure recovers parameters describing absorption and scattering; the concentrations of oxy- and deoxy-haemoglobin and hence StO₂ and the scattering parameters β and α describing the exponential dependence of scattering on wavelength. To demonstrate BB-SRS, a broadband spectrum (700 - 1000 nm, step size 2 nm) was simulated in NIRFAST and was analysed with BB-SRS, SRS and BF. The developed BB-SRS algorithm recovered StO₂ with a relative error of -9%; the concentration of deoxyhaemoglobin with a relative error of +4% , oxyhaemoglobin -10%. The scattering parameters β and α were recovered with a relative error of -30% and -2%, respectively. Among the three algorithms, BB-SRS performed with the best relative error.

In an effort to address Monte Carlo (MC) light prorogation shortcomings in terms of computational burden and light polarization-sensitivity, we report an efficient GPU-based MC for modeling polarized light in scattering medium.

About one in three births in the United States is through Cesarean section. Current monitoring techniques are insufficient to determine hypoxia and acidosis in the fetus during labor. An FDA approved transvaginal fetal pulse oximeter has been used in clinical trials to show that the device can help decrease the rate of Cesarean section. However, this technique has not been adapted into normal hospital procedure. Past pre-clinical and clinical studies have shown the feasibility of transabdominal fetal pulse oximetry. To understand the fundamentals of transabominal fetal pulse oximetry, we examined a layer model with both Monte Carlo and NIRFAST simulations. The NIRFAST model was used to model concentric spheres to understand the effect on geometry. The simulations were used in order to determine how much optical power can be detected from the fetus with a light source at 850 nm. The signal decreased as the fetal depth increased and as source-detector distance increased. The results can be used to aid in the design of a transabdominal fetal pulse oximeter.

The partial optical path length and spatial sensitivity profile in the head model are predicted by a Monte Carlo simulation to analyze the influence of the hemodynamics in the scalp on the time-gated NIRS imaging using null source-detector separation. The simple subtraction of the NIRS signal detected within the early gate time from that within the late gate time is effective to remove the influence of the hemodynamic signal from the scalp in the brain function image.

In this paper we present a Time Resolved Near Infrared device for bed-side neuromonitoring of ischemic stroke patients. This system features three wavelengths allowing a better and robust retrieval of the absolute values of oxy and deoxyhaemoglobin. The device has been fully characterized following the guidelines of the MEDPHOT and BIP protocols, developed under NEUROPt project. Time Resolved spectroscopy is a promising technology that can provide reproducible results in terms of absorption and scattering coefficients. This portable and non-invasive system has been proven suitable for operation in clinical settings.
Data were collected from a cohort of 47 ischemic stroke patients and, according to their cerebral impairment, compared with normal values obtained from a group of 35 healthy subjects. Significant differences in haemoglobin species concentration and saturation were found between healthy and ischemic stroke patients. In the ischemic area of both recanalized and non-recanalized ischemic stroke patients, deoxy-haemoglobin and total haemoglobin values are higher than in controls, while tissue oxygen saturation values are lower only in recanalized patients.

The short- (i.e. over minutes) and mid-term (i.e. over days/weeks) reproducibility of cerebral tissue saturation measured by time domain-NIRS (TD-NIRS) is investigated in this work. We present a pilot study that assess the reproducibility of the measurement of the tissue saturation (StO₂) of the prefrontal cortex on a sample of 7 healthy adult volunteers. TDNIRS measurements were acquired at 16 wavelengths, from 780 to 870 nm, in steps of 6 nm, and were fitted with the diffusion model for semi-infinite homogenous media. Then the absolute concentration of oxy- and deoxyhemoglobin ([HbO₂], [HHb]) were calculated using Beer-Lambert’s law, in order to calculate the volunteer’s brain tissue saturation (StO₂ = [HbO₂]/([HbO₂]+[HHb])). Three measurement sessions were performed on three different days (within a week interval) to evaluate the mid-term reproducibility of the StO₂. For each session, three measurements were taken (within 10 minutes, with repositioning of the probes) to evaluate the short-term reproducibility of the StO₂. The reproducibility was expressed as the within-volunteer standard deviation (SD_w), calculated using a one-way repeated-measure ANOVA. The SD_w in session 1, 2 and 3 were 0.89%, 1.10% and 0.82% respectively, showing a good short-term reproducibility, and the SD_w for all 3 sessions was 1.43%, showing a good mid-term reproducibility without significant variations in the StO₂ between the 3 sessions. Moreover, the mean ± SD global values of StO₂ over all the measurements (n=63) is 62.8+/- 4.2% which is close to the values reported in the literature for adults.

The feasibility of in utero measurement of cerebral blood flow diffuse correlation spectroscopy was demonstrated in lamb fetuses and compared with measurements outside the uterus.

Near-infrared spectroscopy (NIRS) is a non-invasive optical technique that is sensitive to blood volume, blood flow, and oxygen consumption in biological tissue. In particular, a NIRS-measured quantity that has been previously considered as a surrogate for blood flow measurements is the difference of oxy- and deoxy-hemoglobin concentrations ([HbD] = [HbO₂] – [Hb]). We propose a new NIRS method for measurements of cerebral blood flow (CBF), which improves on the [HbD] surrogate by accounting for blood volume contributions and for temporal delays due to the blood transit time in the microvasculature. This new NIRS method relies on concepts of coherent hemodynamics spectroscopy (CHS), and we identify it with the acronym NIRS-CHS. We report a comparison of CBF transient dynamics measured on human subjects with NIRS-CHS, with the [HbD] surrogate and with diffuse correlation spectroscopy (DCS). We found a good agreement between the CBF dynamics measured with NIRS-CHS and with DCS, while the [HbD] dynamics lag because of the delayed effect of CBF on [HbD] due to the capillary and venous blood transit times. The NIRS-CHS method also affords absolute measurements of baseline CBF, for which we found a value of 69 ± 6 ml/100g/min (mean ± standard error) in a group of six healthy volunteers. Further studies to characterize and validate CBF measurements with NIRS-CHS are currently ongoing, with an emphasis on the assessment of accuracy, precision, and reproducibility.

Measuring intracranial pressure (ICP) is typically a highly invasive procedure, in which a ventricular catheter or pressure sensor is placed into the brain. To improve the availability of ICP measurements in non-intensive care patients and research and to reduce the invasiveness and underlying risks of ICP sensing, we developed a non-invasive method to measure ICP with Diffuse Correlation Spectroscopy (DCS) and machine learning. ICP baseline changes were induced in non-human primates (Macaca mulatta) through adjusting the height of a saline reservoir connected to the lateral ventricle via a catheter. ICP was precisely measured with an invasive parenchymal pressure sensor. Cerebral blood flow (CBF) was measured with DCS. The DCS system was operated by a software correlator able to resolve cardiac pulse waves at a sampling rate of 100Hz. To increase signal-to-noise ratio, multiple cardiac pulse waves in CBF were averaged based on systolic peak maximum in invasively measured arterial blood pressure. We hypothesized that the cerebral blood flow pulse waves will change their shape with increasing ICP. The shape of the curve was expressed in numerical features and passed into a regression forest training algorithm. Preliminary results show successful prediction of underlying ICP baselines by the decision forest in one animal. The prediction of non-invasive ICP was achieved with a sampling rate of 1 Hz, an equivalent of about 120 averaged pulses. A larger data set for increased generalizability is the next step to push this approach further.

We present the current status of the LUCA project whose aim is to develop an innovative device combining ultrasound and diffuse optics for an improved screening of the thyroid cancer.

We propose a time domain speckle contrast optical spectroscopy (SCOS) system that makes use of a gated detector and pulsed light source to measure the blood flow variations at very short, quasi-null (<3mm) source-detector separation. We present the results of a human arm cuff occlusion and a comparison with standard SCOS, highlighting that we can probe deeper into tissue, reduce probe footprint, make efficient use of the signal and decrease cost.3

Bioluminescent Imaging (BLI) is a widely utilised technique for the investigation of biological functions within preclinical biomedical studies. Its aim is to image distributed (biologically informative) visible and near infrared light sources, such as luciferase-tagged cells that are located within a living animal. Images are used to estimate the concentration and spatial distribution of reporters and therefore infer biological activity from measurements taken at the surface of the animal. Quantitative accuracy of the measurements can be improved by considering the highly attenuating and scattering nature of biological tissue, modelling the transport of the light through tissue to tomographically reconstruct a 3D image of the light source within the animal. This accuracy can be improved further by collecting spectral data of the bioluminescent signal. Compressive Sensing (CS) is a method of signal processing that utilises the sparse nature of real-world signals in order for them to be compressed in some domain. This in turn means that a sparse signal of length n can be represented by k<<n nonzero coefficients with high accuracy. Due to the localisation of bioluminescent sources, which are in sparse in nature, measurements can be collected using a CS based method. This work introduces the development of a CS based hyperspectral bioluminescent imaging system that can be used to collect compressed hyperspectral fluence data of an internal light source at the surface of an animal model. Effects of the number of measurements collected on image reconstruction quality are also investigated.

A near-infrared (NIR) hyperspectral imaging (HSI) system has been developed to measure the hemodynamic (changes in concentration of oxyhemoglobin and deoxyhemoglobin) and the metabolic (changes in concentration of oxidised cytochrome-c-oxidase) responses in the exposed cortex of small animals. Using the extended spectral information of multiple wavelengths in the NIR range between 780 and 900 nm optimal differentiation between the optical signatures of the chromophores (hemoglobin and cytochrome-c-oxidase) can be achieved. The system, called hNIR, is composed of: (1) a high-frame rate, large-format scientific CMOS (sCMOS) camera for image acquisition and (2) a broadband source coupled with a Pellin-Broca prism mounted on a rotating motor for sequential, fast-rate illumination of the target at different spectral bands. The system characterisation highlights the capability of the setup to achieve high spatial resolution over a ~1x1 mm field of view (FOV). Hyperspectral data analysis also includes simulations using a Monte Carlo optical model of HSI, to estimate the average photon pathlength and improve image reconstruction and quantification. The hNIR system described here is an improvement over a previously tested commercial snapshot HSI solution both in terms of spatial resolution and signal-to-noise ratio (SNR). This setup will be used to monitor brain hemodynamic and metabolic changes in the exposed cortex of mice during systemic oxygenation changes.

This is the first multimodal study of cerebral tissue metabolism and perfusion post-hypoxic-ischaemic (HI) brain injury with broadband near-infrared spectroscopy (bNIRS), diffuse correlation spectroscopy (DCS), positron emission tomography (PET) and magnetic resonance spectroscopy (MRS). In 5 piglet models of HI, we measured cerebral tissue saturation (StO₂), cerebral blood flow (CBF), cerebral oxygen metabolism (CMRO₂), changes in the mitochondrial oxidation state of cytochrome-c-oxidase (oxCCO), cerebral glucose metabolism (CMRglc), and tissue biochemistry (Lac+Thr/tNAA). At baseline, the parameters measured were: 64±6 % StO₂, 35±11 ml/100g/min CBF, and 2.0±0.4 μmol/100g/min CMRO₂. After HI the parameters measured were: 68±6% StO₂, 35±6 ml/100g/min CBF, 1.3±0.1 μmol/100g/min CMRO₂, 0.4±0.2 Lac+Thr/tNAA, and 9.5±2.0 CMRglc. This study demonstrates the capacity of a multimodal set up to interrogate the pathophysiology of HIE using a combination of optical methods, MRS, and PET.

We developed a combined near-infrared reflectance and transmittance approach to enable depth-selective characterization of renal hemodynamics and oxygenation in rats. Test interventions such as arterial occlusion reveal differences in the response between renal cortex and medulla.

We developed an algorithm to estimate the optical properties of a bilayer material using diffuse reflectance analysis. This algorithm has been tested to the detection of liveness in a biometric device adapted to perform Structured Light Imaging. The liveness detection is based on the optical properties comparison between spoofs and living objects.

The luminescence spectra of upconversion nanoparticles (UCNPs) and functionalized UCNPs (UCNPs+folic acid) delivered via subcutaneous and intratumor injection into rats were measured. The particles in normal skin do not diffuse and dissolve, but remain at the introduction site. Inflammatory reaction is developing in the skin when UCNPs were injected. The dense connective tissue capsule is formed around the particles in 3 days after administration. UCNPs were located in the tumor diffusely after 4 daysof injection. Up to 80-90% of the tumor was necrotized after laser irradiation on the fourth day after UCNP administration.

Performance assessment of instruments is a growing demand in the diffuse optics community and there is a definite need to get together to address this issue. Within the EU Network BITMAP¹, we initiated a campaign for the performance evaluation of 10 diffuse optical instrumentation from 7 partner institutions adopting a set of 3 well accepted, standardized protocols. A preliminary analysis of the outcome along with future perspectives will be presented.

Open Data philosophy is becoming more popular among scientists. Open Data approach aims to transform science by making high-quality and well-documented scientific data open to everybody in order to promote collaboration and transparency. In diffuse optical and near-infrared spectroscopy community, a large measurement dataset collected with state-of-the-art instrumentation applied on well-defined phantoms is still missing. Within that context, several European labs from BitMap network¹ have collected diffuse optical data on standard phantoms involving the largest set of diffuse optics instruments published until now. In this work, we present a running project on the open dataset and associated reporting tools.

We present a well-tested, broadband (600-1100 nm) characterized phantom recipe to manufacture tissue mimicking optical phantoms over a wider range of optical properties (absorption 0.1-1 cm-1, reduced scattering 5-25 cm-1) relevant to human organs. The results of various tests like linearity, reproducibility, homogeneity showed the phantom recipe is robust with less than 4 % coefficient of variation (CV). Finally, a non-scattering 3D phantom of the infant's torso was presented to project the futuristic aspect of our work that is to 3D print human organs of biomedical relevance.

The capability of three different tissue oximeters exploiting visible and near-infrared light to determine oxy- and deoxyhemoglobin concentrations and the blood oxygen saturation has been tested in a blood-lipid liquid phantom.

Extracting pathology information embedded within surface optical properties in Spatial Frequency Domain Imaging (SFDI) datasets is still a rather cumbersome nonlinear translation problem, mainly constrained by intrasample and interpatient variability, as well as dataset size. The β-variational autoencoder (β-VAE) is a rather novel dimensionality reduction technique where a tractable set of latent low-dimensional embeddings can be obtained from a given dataset. These embeddings can then be sampled to synthesize new data, providing further insight into pathology variability as well as differentiability in terms of optical properties. Its applications for data classification and breast margin delineation are also discussed.

Rheumatoid arthritis (RA) is an auto-inflammatory disease that causes pain, swelling and stiffness in joints. Diffuse optical tomography (DOT) has shown promise as a non-invasive, diagnostic imaging tool for RA. However high intersubject variability of derived optical parameters infer that at an early stage, small pathophysiological changes resulting from inflammation may be difficult to detect. A set of deep neural network models for RA classification is proposed together with a numerical model of the finger to generate data to overcome the inherent problem of insufficient clinical DOT images available. These proposed deep neural network models have been applied to automatically classify DOT images of inflamed and non-inflamed joints. The results demonstrate that three proposed deep neural network methods improve the diagnostic accuracy as compared to the widely applied support vector machine, especially for high intersubject variability cases. Residual network achieved the highest accuracy (>99%) on the generated database, and highway and convolutional neural networks reached 99% and 90%, respectively. The results show that deep neural network methods are highly suitable for RA classification from DOT data and highlight the potential for deep neural network methods to be used as a computer aided tool in DOT diagnostic systems.

We present results of clinical studies in patients during increase in intra-abdominal pressure (IAP). Changes in brain hemoglobin concentration assessed from time-resolved nearinfrared spectroscopy system were analyzed in frequency domain. The amplitude of power spectral density in respiratory band increases while IAP increases what is related to reduced venous outflow.

Glaucoma is a multifactorial optic neuropathy characterized by progressive loss of retinal ganglion cells, changes in optic disk morphology and visual field defects; its pathophysiology is still unclear. Recently it was demonstrated that glaucoma can be associated with a degenerative effect at the level of the optic nerve and the primary visual cortex. Functional near infrared spectroscopy (fNIRS) is a non-invasive optical technique, which allows the brain hemodynamic monitoring. In particular, the Time Domain fNIRS (TD-fNIRS) allows to remove from the detected signal the contribution coming from the surface (scalp, skull and cerebral fluid) in order to obtain the brain hemodynamic activation. The aim of this preliminary study is to understand if in the glaucomatous patients, the visual cortex activation during a visual stimulus is different from the one of a control group. A total of 20 subjects took part to the study. We divided them into three groups: 7 controls, 5 ocular hypertension (HYPER), and 8 glaucoma. The hemodynamic time courses of oxy- (OHB) and deoxy- (HHB) hemoglobin were compared with a hemodynamic response function (HRF) with the adaptive HRF approach. Finally, an inference test was applied (t-student) to statistically determine the visual cortex activation (simultaneous increase in OHB and decrease in HHB). The p-value threshold was set at 0.05. The 86% of the controls and the 80% of the HYPER combinations are activated; while the 81% of the glaucoma ones are not, outlining a well-defined trend. Also the OHB and HHB show drastic differences between controls and patients.

We localized the thyroid nodules in eleven subjects by ultrasound and measured the microvascular blood flow of the nodules by diffuse correlation spectroscopy.

We show the potential of Diffuse Reflectance Spectroscopy to identify Onchocerca Volvulus nodules, cysts, ganglions and lipomas on humans. We develop a new probe to take into account the geometry of the O. Volvulus model and implement a new measurement protocol and analysis method to deal with the melanin skin contribution. The nodule has a specific intensity and spectral signature compared to its background. It is therefore a relative measure that estimates a signal variation due to the worm and its environment. This may be used to monitor the worm state along time.

The paper presents the results of extended spectral analysis of bone tissue in osteoporosis and evaluation of the effectiveness of correction of this pathological condition using allogeneic hydroxyapatite (HAP). It is shown that the optical method allows to determine the most significant criteria for evaluating the effectiveness of correction using allogeneic hydroxyapatite of different concentrations.

The goal of our study is to train artificial neural networks (ANN) using multispectral images of melanoma. Since the number of multispectral images of melanomas is limited, we offer to synthesize them from multispectral images of benign skin lesions. We used the previously created melanoma diagnostic criterion p'. This criterion is calculated from multispectral images of skin lesions captured under 526nm, 663nm, and 964nm LED illumination. We synthesize these three images from multispectral images of nevus so that the p' map matches the melanoma criteria (the values in the lesion area is >1, respectively). Demonstrated results show that by transforming multispectral images of benign nevus is possible to get a reliable multispectral images of melanoma usable for ANN training.

Monte-Carlo calculations are carried out to simulate the light transport in dense materials. Focus lies on the calculation of diffuse light transmission through films of scattering and absorbing media considering additionally the effect of dependent scattering. Different influences like interaction type between particles, particle size, composition etc. can be studied by this program. Simulations in this study show major influences on the diffuse transmission. Further simulations are carried out to model a sunscreen film and study best compositions of this film and will be presented.

This paper presents the results of the experimental study of the use of ultrasonic treatment for the production of bone implants using Raman spectroscopy (RS). It shows that Raman spectroscopy can be used to estimate a change of bone implants composition during their processing. It was shown that ultrasonic treatment reduces the concentration of cellular components in the bone tissue. The conclusion was based on the two-dimensional analysis of the given ratios.

This paper presents a spectral analysis of various zones of the surfaces of bone and cartilaginous regenerates using the Raman spectroscopy method combined with surface scanning using a stepper motor. It was established that the regenerates at the studied stages of development and the cartilage zone have a similar composition of the relative content of GAG and amides. It indicates a good regenerative potentia.

Animal models present a specific set of challenges for hyperspectral imaging. Sample fixation using formalin changes the optical properties of tissues. Since tissues in murine models can be thin and translucent, substrate selection also plays a key role in properly setting up an experimental protocol.

Diffuse reflectance spectroscopy was successfully applied in transfusion medicine, for the diagnosis of the quality of blood units, in recent studies. However, removal of blood sample from the bag is still required for measuring donor dependent hematological variations in stored units. If these hematological variables are successfully interpreted from diffuse reflectance spectra, there would be no need for invasive tests. In this study, diffuse reflectance spectra of 103 erythrocyte suspensions were modeled with their donor-dependent hematological variations on the first day of storage. The final model given in this study, fitted well at 610 nm with goodness of fit of 0.94 for erythrocyte suspensions and 0.96 for 87 leukocyte-depleted units, when selecting 474 nm for normalizing diffuse reflectance spectra of each blood unit and minimizing the effects of plastic blood bag.

The proposed method relates to spectroscopic ones and is based on the measurement of the intensity of the absorbed irradiation when it passes through the blood-containing organ of a patient. This way of non-invasive determination of glucose contents in human blood was proposed far ago but the reason of its absence in the market is the non-solved problem of the irradiation part definition which is absorbed just by glucose but not by the other absorbing components of the human organ. This problem solution includes measurements of irradiation absorption at numerous wavelengths of irradiation and application of specific mathematical apparatus which we propose to apply for this method. The mathematical calculations include the solution of linear equations system with light intensity parts which are absorbed by different absorbing components as variables. The number of equations id equal to the amount of absorbing components and the number of wavelengths at which the measurements must be performed. The calculations allowing to transfer the part of absorbed light into glucose concentration also are listed

Skin spectral reflectance can be used for estimate optical properties of human skin tissue, but there is a potential problem. We investigated skin conditions producing similar reflectance spectra but different point spread functions of reflected intensity on the skin surface in the framework of Monte Carlo simulation.

Skin monitoring under different diffuse spectral illumination conditions is important for assessment of a number of skin, respiratory and circulation conditions. We developed a prototype portable device designed for reflectivity of full back of hand with irradiation with blue LEDs. We are currently applying it to monitoring of vitiligo and age spots, and their temporal evolution.

Monte Carlo (MC) method is regarded as the gold standard for modeling the light transport in biological tissues. Due to the stochastic nature of the MC method, many photon packets need to be processed to obtain an adequate quality of the simulated reflectance. The number of required photon packets further increases if the numerical aperture of the detection scheme is low. Consequently, extensively long simulation times may be required to obtain adequate quality of the reflectance for such detection schemes. In this paper we propose an efficient regression model that maps reflectance simulated at the maximum acceptance angle of 90◦ to the reflectance corresponding to a much smaller realistic acceptance angle. The results of validation on spatially resolved reflectance and inverse models for estimation of optical properties show that the regression models are accurate and do not introduce additional errors into the spatially resolved reflectance or the optical properties estimated by appropriate inverse models from the regressed reflectance.

We present a diffuse optical imaging system with structured illumination and integrated detection for spatial characterization of scattering and absorption properties of turbid media. It is based on the application of single- pixel imaging techniques with integrating spheres, which allows us to develop a spatial resolved version of the Kubelka-Munk method.

Functional near-infrared spectroscopy as an optical (i.e., light-based) neuroimaging technique is susceptive to ambient light noise. In the daily task scenario experiments, light is required to monitor the movement of patients and to minimize the effect of this light on the results; the fNIRS optodes are covered with dark materials (e.g., a loose-fitting black shower-cap). In our laboratory, over-caps provided by NIRx (produced by EasyCap) have been used to eliminate this con-founder. However, there has been a demand to evaluate their performance by quantification. Thus, in this paper, the transmission of light through a fabric over-cap is investigated. The results revealed that the output signal of functional near-infrared spectroscopy could be contaminated significantly by the ambient light. Moreover, the noise varies due to the stretch that could be applied due to various head sizes. The changes in the amplitude of the signal, which relates to physiological responses, is about 1-2 % in continuous wave measurement while the result of the investigation notes that the transmission average in the samples varies from 8-46 percent depending on the stretching forces. Therefore, it is suggested that this kind of over-caps would be applied only in a dimmed environment, which is not applicable when subjects have mobility disorders. Under such conditions, other techniques to minimize ambient light should be practiced.

It has been reported that the use of illumination with sinusoidal intensity pattern is efficient in noncontact optical tomography. Although this imaging method of using spatial frequency is promising, most research so far has relied on the diffusion approximation to the radiative transport equation. Here, a numerical algorithm of optical tomography with spatially sinusoidal illumination for the radiative transport equation are proposed. With the help of the technique of rotated reference frames, the forward problem, i.e., the three-dimensional radiative transport equation is solved by the three-dimensional FN method. Then the inverse problem is solved by making use of the Green’s function, which is the solution to the forward problem.

Absolute concentrations of oxyhaemoglobin and deoxyhaemoglobin obtained using multi-wavelength measurements of time-resolved diffuse reflectance signals are presented. The aim was to test how accurately the concentration of oxy- and deoxyhemoglobin can be measured simultaneously at multiple wavelengths. The optical signals were collected using the system constructed by the author’s group^a , which records the distribution in time of flight of photons (DTOFs) simultaneously at 16 spectral channels ranging from 650 to 850 nm. The measurements were carried out on liquid phantoms containing intralipid solution, human blood and yeast in varying amounts. The oxygen saturation of blood covered a continuous range from 100 % to 0 % during 6 deoxygenation cycles. The estimated values of total haemoglobin (20.9, 35.7, 57.3, 45.7 μM) are close to the reference values obtained using a blood gas analyzer (21.3, 37.0, 57.3, 48.0 μM) and the estimated values of absolute concentrations of oxy-, deoxyand total haemoglobin are similar to the values obtained using a commercial frequency-domain NIRS system (OxiplexTS ^TM ). The phantom measurements have shown the capability to measure the absolute concentrations of chromophores in a studied media using multi-wavelength, time-resolved NIRS technique. The excess number of spectral channels can potentially be used to resolve changes in oxidation state of cytochrome-c-oxidase enzyme.

The spectral study results and comparative evaluation of the component composition of the surfaces of hard palate implant samples made by the technology "Lioplast", used in the clinic in the field of atrophic processes in multiple recessions of the gums, using the method of Raman spectroscopy (RS) are presented. It is shown that Raman spectroscopy can be used to assess changes in the composition of implants based on the hard palate during their processing. The coefficients were introduced and a two-dimensional analysis was carried out, which showed that the main components are preserved and DNA/RNA is removed during processing, which improves the quality of the material that provides the possibility of a good clinical effect in the treatment of multiple recessions of the gums.

Functional near infrared spectroscopy (NIRS) is a widespread non-invasive technique to monitor skeletal muscle metabolism. However, only variation of oxygenated (HHb), deoxygenated (O₂Hb), total (tHb) hemoglobin and saturation (SO₂) are usually reported. In this study, Time Domain (TD) NIRS approach was exploited to perform a preliminary quantitative characterization of vastus lateralis muscle during incremental exercise. A population of 11 healthy young male subjects performed on a mechanical cycle ergometer an incremental exercise (initial work rate range = 60-96 W, increment = 12-18 W/min) until exhaustion. TD NIRS, heart rate, pulmonary ventilation (VE), O₂ uptake (VO₂), CO₂ output (VCO₂), blood lactate concentration ([La]b) and Borg scale were measured during the exercise. From TD NIRS, muscles absolute values of absorption and scattering coefficients were obtained with a homogeneous approach and hemoglobin concentrations and saturation levels were calculated. The time courses of HHb, O₂Hb, tHb and SO₂ were consistent with previous literature results. A high inter-subject variability was found for both optical properties and hemodynamic concentrations. Further statistical group analysis will be required in order to highlight significant behavior within the population and correlation with physiological parameters.

Epilepsy is a neurological disorder characterized by chronic excessive neuronal discharges. Epilepsy surgery may be considered in patients who are resistant to drug treatment, and several tests are used during pre-operative planning to locate the epileptic focus to be resected (MRI, PET, SPECT, EEG, MEG). In some cases, intracranial EEG monitoring or intraoperative electrocorticography is required to confirm and better delineate the area to be resected. Despite available tests, epilepsy surgery outcome remains modest. We present a multimodal imaging platform connected to a neurosurgical microscope allowing intraoperative detection of intrinsic optical brain biomarkers for surgical guidance during epilepsy surgery. Hyperspectral imaging (HSI) allows detection of the hemodynamic response associated with epileptic activity, spatial frequency domain imaging (SFDI) is used for optical properties reconstruction (absorption and reduced scattering), and auto-fluorescence imaging (AFI) allows metabolic markers identification. Validation of SFDI and AFI systems was performed on optical phantoms. Acquisitions on ex vivo tissue samples demonstrate the capabilities of the system to produce fluorescence intensity maps, calibrated with optical properties obtained from SFDI to account for tissue attenuation. In vivo acquisition during epilepsy surgery with HSI allows characterization of the hemodynamic response associated with epileptic spikes.

Time Domain Diffuse Optical Tomography (TD-DOT) at different wavelengths can be used to retrieve tissue reconstructing the components of a two-region system starting from self-normalized time-dependent measurements performed in reflectivity geometry over multiple wavelengths. The proposed method performs a fit of a limited number of tissues parameters providing a good quantification of the components’ concentrations by applying a FEM-based Diffusion approximation of the TD-DOT direct model.

Blood oxygenation in depth is a critical parameter to be monitored in many clinical applications. We are developing a pre-clinical instrument to monitor oxygenation up to 1.5 cm deep in biological tissues. The measures, based on time of ight, are in vivo, non invasive and without injection of any contrast agent. Developments have been implemented in a tomographic reconstruction algorithm to make it reconstruct chromophores by layers. These developments are presented here along with simulations we used to set up the algorithm. Buried aps on 20 pigs were undergone for pre-clinical tests of our reconstruction algorithm and instrument.

We have studied experimentally and theoretically spatial distributions of factors describing sensitivity of the statistical moments of distributions of times of flight (DTOFs) of photons penetrating through the medium to changes in absorption coefficient. Additionally, the moments subtraction procedure, based on difference between statistical moments measured at two interoptode distances was applied in order to modify the sensitivity profiles.

In this paper, we propose a novel calibration procedure based on modeling and measurement of the reflectance distance profiles from a metallic mirror. We observe a remarkable agreement of our reflectance distance profile model with the measurements yielding repeatable calibration factors within 2% when tested on silver and aluminum mirrors. Comparison to widely acceptable calibration using polystyrene microspheres suspensions yields errors of below 10%.

An apparatus for direct determination of the spectrally resolved scattering phase function of suspensions and emulsions is described. The system has a polychromatic xenon lamp as light source and a spectrometer as detector. In combination with a stepper motor the system enables sepctrally and angularly resolved measurements in the range of 450nm to 950nm and 10° to 170°, respectively. A post processing algorithm, which takes the light propagation inside the cuvette within the regime of single scattering and the spectral dimension into account, was developed. This allows a direct determination of the scattering phase function for the indicated spectral and angular range. By comparing measurements on polystyrene microspheres with Mie theory the concept of the presented instrument was validated. Finally, the method was used to determine the scattering phase functions of different types and brands of soybean oil emulsions, a common phantom material in the field of diffuse optics.

We propose a nonlocal diffusion equation (NDE) as a new forward model, which uses the concepts of differential operators under the nonlocal vector calculus. The discretization of the NDE is performed using an effective graph-based numerical method (GNM). We evaluate the proposed forward modelling method on a homogeneous slab where the analytical solution is available. Our experiments show that the results of the NDE (discretized by GNM) is quantitatively comparable to the analytical solution. The proposed method has an identical implementation for geometries in two and three dimensions due to the nature of the graph representation.

A continuous wave broadband near-infrared spectroscopy method is developed based on twolayered fitting of optical properties, to recover the haemoglobin concentrations and scattering parameters in cerebral and extra-cerebral tissues along with fitting for the extra-cerebral layer thickness. It is shown that tissue oxygenation for deep tissue and superficial tissue thickness is recovered with less than 4% error, whereas a homogeneous fit having an error of 15%.

Curvature and height corrections were made on hyperspectral images in order to reduce intensity dependence on distance and inclination. The corrections were made by Lambertian cosine law and measured 3D surface of the sample. Image of homogeneous phantom appeared significantly more homogeneous after correction.

Development of cost-effective methods and instruments to accurately determine optical parameters of turbid materials can have wide applications in life science research and clinical applications such as blood test and assay of cell suspensions. We report a fast and photodiode based method without integrating spheres to measure three light scattering signals and determine absorption coefficient, scattering coefficient and anisotropy factor of the sample using only one optically thick sample. Results of optical parameters for microsphere suspension samples have been obtained from 460 to 1000nm and validated against the results predicted by Mie theory. Using a GPU executed Monte Carlo code and gradient decent based inverse algorithm, the inverse solutions for determination of optical parameters and their spectra can be achieved in real-time. The uniqueness of the inverse solutions has been proved.

Pulse oximeter has been a widely used instrument in hospitals or household for measuring arterial oxygen saturation value. Its annual global market value has exceeded USD 1 billion and is still growing. There are two main shortcomings in most pulse oximeters on the market: 1. database based algorithm limits the range of precision, and 2. incapable of measuring tissue oxygen saturation related to local tissue oxygen exchange condition. We developed a 3-wavelength frequency multiplexed intensity modulated diffuse reflectance system in which the modulation frequencies of the 3 light sources are different so that the light reflectance at the 3 wavelengths can be collected by the detector simultaneously. This allows us achieving determination of absorption and reduced scattering coefficients at the 3 wavelengths of the exactly same instant of time and this can lead to accurate quantification of tissue and arterial oxygen saturations. This technique works in conjunction with a photon diffusion equation and does not need to build a database in advance to calculate arterial oxygen saturation. In addition, capability of this system in measuring tissue oxygen saturation would facilitate swift monitoring of local tissue blood circulation.

Diffuse Optical Tomography (DOT) is a tool for 3D reconstruction of absorption and scattering inside a tissue. Typically, this method requires a dense distribution of sources and detectors, thus hampering the possibility of fully exploring a time-resolved detection. Recently, techniques based on structured-light illumination and compressing detection have been developed, opening the possibility of fully exploiting a source/detector spatial modulation for compression at the measurement stage. Here we propose a combined Continuous-Wave (CW) and time-domain (TD) adaptive scheme based on the singular-value decomposition (SVD) for optimal-patterns calculation. Patterns are firstly computed based on a fast acquisition via a CCD, and consequently projected for time-resolved measurements.

Optical methods are attractive tools for neuromonitoring given their safety and sensitivity to key markers of brain health: tissue oxygenation can be assessed by near-infrared spectroscopy (NIRS) and cerebral blood flow by diffuse correlation spectroscopy (DCS). Although the application of these tools to neonatal patients is fairly straightforward, since it is reasonable to model the head as an optically homogeneous medium, their use with adult patients is more complicated due to substantial signal contamination caused by hemodynamic fluctuations in the extracerebral (EC) tissue. The purpose of this study was to assess the magnitude of this contamination by acquiring NIRS and DCS data in response to a hypercapnic challenge with and without scalp contributions. Scalp blood flow was impeded by a pneumatic tourniquet, which was confirmed by dynamic contrast-enhanced (DCE) NIRS. The results showed that EC contamination for intensity measurements could be as high as 75%; however, using time-resolved detection can reduce this value to 30%.

We evaluate a method to obtain the pathlength distribution of detected photons from reflectance as a function of absorption coefficient through inverse Laplace transformation using curve fitting as a noise reduction technique. We numerically validate our method using Monte Carlo simulations and exact solutions of time-resolved and steady state diffuse reflectance. Knowledge of the path length distribution may be helpful in drafting models for sub-diffuse reflectance measurements such as Single Fiber Reflectance Spectroscopy, whereas moments of the pathlength distribution (mean and variance) may provide diagnostic information by themselves.

The changes in the oxy- and deoxygenated hemoglobins at different depths of skin tissue were obtained from the diffuse reflectance spectra in three wavelength ranges of the skin. The influence of the optical path length in the calculation on the results of the hemoglobin change was evaluated.

Radiofrequency ablation (RFA) is minimally invasive thermotherapy, where a heating source is used to target and kill malignant cells in a tissue. While RFA has tremendous potential in the field of oncology, there is also a need for reliable real-time monitoring of this procedure to avoid over or under treatment. In this work, we investigate the use of timeresolved diffuse optical spectroscopy (DOS) to continuously track the change in optical properties during RFA to monitor the process of ablation. The time evolution of the spectra of the optical properties of the tissue undergoing treatment gives deep insights into the structural and constitutional changes occurring during the RFA treatment.

We introduce an implementation of the Extended Kalman Filter for the retrieval of absorption coefficients in layered turbid media. The technique was validated with experiments in a liquid phantom.

To measure optically physiological condition inside skin tissue, it is important to estimate optical parameters in skin tissue. In this study, we investigated a method to estimate absorption and scattering parameters in skin tissue from the spectral reflectance database constructed by using Monte Carlo simulation with a nine-layered skin tissue model.

In this paper we present the ex-vivo characterization of a full-custom made multi-wavelength, two channel Time-Resolved Spectroscopy (TRS) module developed with the aim of being integrated in to a multi-modal spectroscopic device. This module overcomes all the main drawbacks of systems based on time-domain techniques such as high complexity and bulkiness while guaranteeing performances comparable to expensive state-of-the-art available devices. Each subcomponent of the module has been tailored and optimized to meet all the above-mentioned requirements. In order to assess and translate the performances of these tools for effective clinical use, we characterized the system following the guidelines of common standardization protocols. By following MEDPHOT guidelines, the linearity and accuracy in retrieving absolute values of absorption and scattering coefficients were determined by means of measurements on homogeneous phantoms. Finally, by means of a mechanically switchable solid inhomogeneous phantom (developed under the nEUROPT project) we simulated the clinical problem of detecting and localizing an absorption perturbation in a homogeneous background with broad applications such as detection of cancer lesions, thyroid, etc.

Time-resolved (TR) techniques are exploited in many biomedical applications in order to find absolute values of absorption (μ_a) and reduced scattering (μ_s’) coefficients that characterize biological tissues chemical and microstructure properties. However, the concomitant acquisition of tissue distribution time-of-flight (DTOF) and instrument response function (IRF) is necessary to perform quantitative measurements. This can be a non-trivial time consuming operation which typically requires to detach the optical fibers from the measurement probe (usually put in a reflectance configuration for in-vivo applications) in order to face them one to each other (“reference” geometry). To overcome these difficulties, a new IRF measurement method that exploit the “reflectance” geometry is here proposed. A practical 3D printed implementation has been carried out for a specific device to test the feasibility of this approach and if the IRF acquired in the “reflectance” geometry is equivalent to the “reference” one. A particular problem addressed is the determination of the temporal shift T0 that can occur between IRF and sample DTOF. Two different approaches, based respectively on the curves barycenters difference and on a calibration phantom, are proposed. Both methods are valid and indifferently applicable according to specific measurement requirements. This allows “reflectance” IRF acquisition to be eligible as standard methodology for TR measurements.

We apply first order perturbation theory to the scalar radiative transport equation for the temporal field autocorrelation function to study DCT and SCOT sensitivity to changes in the Brownian motion of the constituent scattering particles.

Diffuse optical imaging can be used to probe highly scattering media like biological tissue down to a depth of few centimeters, with spatial resolution limited by light scattering. Its combination with ultrasound imaging can potentially lead to medical imaging systems with, for instance, high specificity in the examination of tumors. However, the presence of the ultrasound coupling gel between probe and tissue can have detrimental effects on the accuracy of optical imaging techniques. Here we present an experimental study on the effect of ultrasound coupling fluids on diffuse optical spectroscopy (DOS) and diffuse correlation spectroscopy (DCS). We demonstrate on tissue-mimicking phantoms that the use of standard water-clear gels, providing a direct path for the light from the source to the detection point, can distort optical measurements generating strong underestimation of both the absorption and the reduced scattering coefficients in DOS measurements, as well as underestimation of the Brownian diffusion coefficient in DCS measurements. On the contrary, various turbid fluids demonstrate excellent performance in preventing this issue.

For the first time, we are proposing a compressive-sensing approach to time-domain diffuse Raman spectroscopy for depth probing of multilayer diffusive media. We built a spectrometer capable of both spectral and temporal acquisition with a single-pixel detector and tested it for depth sectioning on a bilayer tissue-mimicking phantom.

Time-domain diffuse correlation spectroscopy (TD-DCS) is an emerging optical technique with the potential to resolve the blood flow (BF) in depth. The first in vivo measurements have been shown recently on humans, however improvements in terms of signal-to-noise ratio (SNR) and depth sensitivity would be beneficial for biological applications. In this contribution, we explore the possibility of in vivo TD-DCS measurements above 1000 nm, and discuss its possible advantages compared to standard wavelengths (i.e. 700-800 nm). In our experimental setup, we exploited a tunable pulsed laser source extended more to the infrared and an InGaAs photomultiplier. Here, we report the results of a cuff occlusion on the forearm of a healthy adult subject at a wavelength of 1000 nm. Compared to the same experiment at standard wavelength (785 nm), the electric-field auto-correlation functions show a slower decay rate during all the experiment (both during and after the occlusion) as expected, suggesting a higher SNR. Even longer wavelengths, for diminishing water absorption, can be obtained through optimization of the laser source and the use of more efficient detectors.

We have successfully fabricated and characterized customized 3D printed optical phantoms that simulate not only physiologically relevant absorption and reduced scattering coefficients but also fluorescence properties similar to the widely applied contrast agent indocyanine green.