Search

We developed a compact, hand-held hyperspectral imaging system for 2D neural network-based visualization of skin chromophores and blood oxygenation. State-of-the-art micro-optic multichannel matrix sensor combined with the tunable Fabry-Perot micro interferometer enables a portable diagnostic device sensitive to the changes of the oxygen saturation as well as the variations of blood volume fraction of human skin. Generalized object-oriented Monte Carlo model is used extensively for the training of an artificial neural network utilized for the hyperspectral image processing. In addition, the results are verified and validated via actual experiments with tissue phantoms and human skin in vivo. The proposed approach enables a tool combining both the speed of an artificial neural network processing and the accuracy and flexibility of advanced Monte Carlo modeling. Finally, the results of the feasibility studies and the experimental tests on biotissue phantoms and healthy volunteers are presented.

Recent advances in the development of ultra-compact semiconductor lasers and technology of printed flexible hybrid electronics have opened broad perspectives for the design of new pulse oximetry and photoplethysmography devices. Conceptual design of optical diagnostic devices requires careful selection of various technical parameters, including spectral range; polarization and intensity of incident light; actual size, geometry, and sensitivity of the detector; and mutual position of the source and detector on the surface of skin. In the current study utilizing a unified Monte Carlo computational tool, we explore the variations in diagnostic volume due to arterial blood pulsation for typical transmitted and back-scattered probing configurations in a human finger. The results of computational studies show that the variations in diagnostic volumes due to arterial pulse wave are notably (up to 45%) different in visible and near-infrared spectral ranges in both transmitted and back-scattered probing geometries. While these variations are acceptable for relative measurements in pulse oximetry and/or photoplethysmography, for absolute measurements, an alignment normalization of diagnostic volume is required and can be done by a computational approach utilized in the framework of the current study.

We explore the propagation of the cylindrical vector beams (CVB) in turbid tissue-like scattering medium in comparison with the conventional Gaussian laser beam. The study of propagation of CVB and Gaussian laser beams in the medium is performed utilizing the unified electric field Monte Carlo model. The implemented Monte Carlo model is a part of a generalized on-line computational tool and utilizes parallel computing, executed on the NVIDIA Graphics Processing Units (GPUs) supporting Compute Unified Device Architecture (CUDA). Using extensive computational studies, we demonstrate that after propagation through the turbid tissue-like scattering medium, the degree of fringe contrast for CVB becomes at least twice higher in comparison to the conventional linearly polarized Gaussian beam. The results of simulations agree with the results of experimental studies. Both experimental and theoretical results suggest that there is a high potential of the application of CVB in the diagnosis of biological tissues.

Path tracing is a well-established technique for photo-realistic rendering to simulate light path transport. This method has been widely adopted in visual effects industries to generate high quality synthetic images requiring a large number of samples and a long computation time. Due to the high cost to produce the final output, intermediate previsualization of path tracing is in high demand from production artists to detect errors in the early stage of rendering. But visualizing intermediate results of path tracing is also challenging since the synthesized image with limited samples or improper sampling usually suffers from distracting noise. The ideal solution would be to provide a highly plausible intermediate result in the early stages of rendering, using a small fraction of samples, and apply a posteriori manner to approximate the ground truth. In this thesis, this issue is addressed by providing several efficient posteriori reconstructions and denoising technique for previsualization of pa-th tracing. Firstly, we address the problem for the recovery of the missing values to construct low rank matrices for incomplete images including missing pixel, missing sub-pixel, and multi-frame scenarios. A novel approach utilizing a convolutional neural network which provides fast precompletion for initializing missing values, and subsequent weighted nuclear norm minimization with a parameter adjustment strategy efficiently recovers missing values even in high frequency details. The result shows better visual quality compared to the recent methods including compressed sensing based reconstruction. Furthermore, to mitigate the computation budgets of our new approac-h, we extend our method by applying a block Toeplitz structure forming a low-rank matrix for pixel recovery, and tensor structure for multi-frame recovery. In this manner, the reconstruction time can be significantly reduced. Besides that, by exploiting temporal coherence of multi-frame with a tensor structure

We introduce a noninvasive diagnostic approach for functional monitoring of blood microflows in capillaries and thermoregulatory vessels within the skin. The measuring system is based on the combined use of laser Doppler flowmetry and skin contact thermometry. The obtained results suggest that monitoring of blood microcirculation during the occlusion, performed in conjunction with the skin temperature measurements in the thermally stabilized medium, has a great potential for quantitative assessment of angiospatic dysfunctions of the peripheral blood vessels. The indices of blood flow reserve and temperature response were measured and used as the primarily parameters of the functional diagnostics of the peripheral vessels of skin. Utilizing these parameters, a simple phenomenological model has been suggested to identify patients with angiospastic violations in the vascular system.

With pulse wave periodic beating taken into account we assess the influence of blood volume and oxygen saturation changes on the measurements of diffuse reflectance spectra of human skin in the visible and NIR spectral range.

According to the International Diabetes Federation, the challenge of early stage diagnosis and treatment effectiveness monitoring in diabetes is currently one of the highest priorities in modern healthcare. The potential of combined measurements of skin fluorescence and blood perfusion by the laser Doppler flowmetry method in diagnostics of low limb diabetes complications was evaluated. Using Monte Carlo probabilistic modeling, the diagnostic volume and depth of the diagnosis were evaluated. The experimental study involved 76 patients with type 2 diabetes mellitus. These patients were divided into two groups depending on the degree of complications. The control group consisted of 48 healthy volunteers. The local thermal stimulation was selected as a stimulus on the blood microcirculation system. The experimental studies have shown that diabetic patients have elevated values of normalized fluorescence amplitudes, as well as a lower perfusion response to local heating. In the group of people with diabetes with trophic ulcers, these parameters also significantly differ from the control and diabetes only groups. Thus, the intensity of skin fluorescence and level of tissue blood perfusion can act as markers for various degrees of complications from the beginning of diabetes to the formation of trophic ulcers.

We consider changes in the volume of blood and oxygen saturation caused by a pulse wave and their influence on the diffuse reflectance spectra in the visible/NIR spectral range. CUDA-based Monte-Carlo model was used for routine simulation of detector depth sensitivity (sampling volume) and skin spectra, and their variations associated with physiological changes in the human skin. The results presented in the form of animated graphs of sampling volume changes for scaling of the parameters of the main human skin layers related to the results of experimental measurements are of particular interest for pulse oximetry, photoplethysmography, Doppler flowmetry, reflectance spectroscopy.

In the framework of further development of a unified computational tool for the needs of biomedical optics, we introduce an electric field Monte Carlo (MC) model for simulation of backscattering of coherent linearly polarized light from a turbid tissue-like scattering medium with a rough surface. We consider the laser speckle patterns formation and the role of surface roughness in the depolarization of linearly polarized light backscattered from the medium. The mutual phase shifts due to the photons' pathlength difference within the medium and due to reflection/refraction on the rough surface of the medium are taken into account. The validation of the model includes the creation of the phantoms of various roughness and optical properties, measurements of co-and cross-polarized components of the backscattered/reflected light, its analysis and extensive computer modeling accelerated by parallel computing on the NVIDIA graphics processing units using compute unified device architecture (CUDA). The analysis of the spatial intensity distribution is based on second-order statistics that shows a strong correlation with the surface roughness, both with the results of modeling and experiment. The results of modeling show a good agreement with the results of experimental measurements on phantoms mimicking human skin. The developed MC approach can be used for the direct simulation of light scattered by the turbid scattering medium with various roughness of the surface.

The interaction of ultrasound and light in biological tissues results in a small amount of the scattered light being shifted relative to the carrier frequency (typically 1 part in 108). We have developed an inherently efficient and low noise quantum memory based technique to selectively absorb these ultrasound tagged photons in a pair of atomic frequency combs, and recover them delayed in time as a photon echo. In this manner we have demonstrated record ultrasoundmodulated sideband-to-carrier discrimination (49dB). Further, we confirm that the technique is compatible with highly scattering samples, and present initial acoustic pulse tracking measurements. This strongly suggests the suitability of the technique for biological tissue imaging.

The incidence of the skin melanoma, the most commonly fatal form of skin cancer, is increasing faster than any other potentially preventable cancer. Clinical practice is currently hampered by the lack of the ability to rapidly screen the functional and morphological properties of tissues. In our previous study we show that the quantification of scattered laser light polarization provides a useful metrics for diagnostics of the malignant melanoma. In this study we exploit whether the image speckle could improve skin cancer diagnostic in comparison with the previously used free-space speckle. The study includes skin phantom measurements and computer modeling. To characterize the depolarization of light we measure the spatial distribution of speckle patterns and analyse their depolarization ratio taken into account radial symmetry. We examine the dependences of depolarization ratio vs. roughness for phantoms which optical properties are of the order of skin lesions. We demonstrate that the variation in bulk optical properties initiates the assessable changes in the depolarization ratio. We show that image speckle differentiates phantoms significantly better than free-space speckle. The results of experimental measurements are compared with the results of Monte Carlo simulation.

The influence of a low-frequency electric field applied to soft biological tissues ex vivo at normal conditions and upon the topical application of optical clearing agents has been studied by optical coherence tomography (OCT). The electro-kinetic response of tissues has been observed and quantitatively evaluated by the double correlation OCT approach, utilizing consistent application of an adaptive Wiener filtering and Fourier domain correlation algorithm. The results show that fluctuations, induced by the electric field within the biological tissues are exponentially increased in time. We demonstrate that in comparison to impedance measurements and the mapping of the temperature profile at the surface of the tissue samples, the double correlation OCT approach is much more sensitive to the changes associated with the tissues' electro-kinetic response. We also found that topical application of the optical clearing agent reduces the tissues' electro-kinetic response and is cooling the tissue, thus reducing the temperature induced by the electric current by a few degrees. We anticipate that dcOCT approach can find a new application in bioelectrical impedance analysis and monitoring of the electric properties of biological tissues, including the resistivity of high water content tissues and its variations.

Within the framework of further development of unified Monte Carlo code for the needs of biomedical optics and biophotonics, we present an approach for modeling of coherent polarized light propagation in highly scattering turbid media, such as biological tissues. The temporal coherence of light, linear and circular polarization, interference, and the helicity flip of circularly polarized light due to reflection at the medium boundary and/ or backscattering events are taken into account. To achieve higher accuracy in the results and to speed up the modeling, the implementation of the code utilizes parallel computing on NVIDIA graphics processing units using Compute Unified Device Architecture. The results of the simulation of coherent linearly and circularly polarized light are presented in comparison with the results of known theoretical studies and the results of alternative modelings.

We exploit the directional awareness of circularly and/or elliptically polarized light propagating within media which exhibit high numbers of scattering events. By tracking the Stokes vector of the detected light on the Poincaŕe sphere, we demonstrate its applicability for characterization of anisotropy of scattering. A phenomenological model is shown to have an excellent agreement with the experimental data and with the results obtained by the polarization tracking Monte Carlo model, developed in house. By analogy to diffusing-wave spectroscopy we call this approach diffusing-wave polarimetry, and illustrate its utility in probing cancerous and non-cancerous tissue samplesin vitro for diagnostic purposes.