Here, we present a technique for inverting a Monte Carlo simulation to extract tissue optical properties from the moments of the spatio-temporal response of the tissue by training a 5-layer fully connected neural network. Initial reports of this model were used to extract optical properties from a single layer tissue model with high accuracy. Here, we expand this method to demonstrate its ability to extract optical properties from individual layers in a multi-layer model. We demonstrate the accuracy of the method across a very wide parameter space and demonstrate that the method is insensitive to parameter selection of the neural network model itself.
Monte Carlo Simulations (MCSs) allow for the estimation of photon propagation through media given knowledge of the geometry and optical properties. Previous research has demonstrated that the inverse of this problem may be solved as well, where neural networks trained on photon distributions can be used to estimate refractive index, scattering and absorption coefficients. To extend this work, time-dependent MCSs are used to generate data sets of photon propagation through various media. These simulations were treated as stacks of 2D images in time and used to train convolutional networks to estimate tissue parameters. To find potential features that drive network performance on this task, networks were randomly generated. Generated networks were then trained. The networks were validated using 4-fold cross validation. The consistently performing top 10 networks typically had an emphasis on convolutional chains and convolutional chains ending in max pooling.
There have been concerted efforts to develop high-energy diode-pumped alkali vapor lasers (DPAL). These hybrid gas phase / solid-state laser systems offer possibilities for constructing high-powered lasers that have high beam quality. Considerable progress has been made, but there are technical challenges associated with the reactivity of the metal atoms. Rare gas atoms (Rg) excited to the np5 (n+1)s 3P2 configuration are metastable and have spectral properties that are closely similar to those of the alkali metals. Optically pumped lasers have been constructed using excitation of the np5 (n+1)p ← np5 (n+1)s transitions. Pulsed lasing has been observed for Ne*, Ar*, Kr* and Xe*. Helium was used as the collisional energy transfer agent that established population inversions. These systems have the advantage using inert reagents that are gases at room temperature, with excellent potential for closed-cycle, multi-wavelength operation. The primary technical difficulty for the rare gas laser is the discharge production of sufficient Rg* metastables in the presence of >200 Torr of He. We have developed a high frequency pulsed discharge that yields >1013 cm-3 Ar* in the presence of He at pressures up to 730 Torr. Using this discharge, a diode pumped Ar* laser providing 4.1 W of continuous wave output has been demonstrated, with an optical conversion efficiency of 31%. Development of the pulsed discharge system and CW lasing demonstrations with Xe* are reported.
Detection of targets can be difficult using thermal camera during thermal cross-over periods. Under these conditions, the target takes on the apparent temperature of objects in the foreground or background and becomes undetectable. It is commonly said that the thermal imagery is “washed-out”. When the thermal contrast of an object against its background is zero, many times a polarization contrast of the same object is non-zero. In this paper, we introduce a camera that measures thermal and polarization images in both the mid-wave infrared (MWIR) and long-wave infrared (LWIR). We also show example images and derive a simple equation that explains the conditions under which a polarization signature of an object can be expected.
A simulation study was conducted for the purpose of identifying technology improvements for an acquisition sensor for the detection of small objects in clear, sunlit cloud, fog, and mist conditions. Currently available mid-wave infrared (MWIR) and long-wave infrared (LWIR) technologies were studied. In addition, projected sensor technologies anticipated to be available in the near future, as well as idealized systems limited only by aperture size, integration time and instantaneous field of view (IFOV) were modeled. Both standard and polarimetric imaging sensors were included in the study. The Aero-Optical Prediction Tool (AerOPT) was used to model the performance of various sensors operating under the conditions of interest. Results indicate that LWIR systems may extend detection range in fog and mist environments and that polarimetry may reduce false alarm rate for sunlit cloud backgrounds. Importantly, polarimetric imaging does not appear to negatively impact detections.
KEYWORDS: Fiber optic gyroscopes, Mie scattering, Light scattering, Scattering, Monte Carlo methods, Atmospheric particles, Photon transport, Atmospheric propagation, Air contamination, Visibility through fog
Anyone who has driven through fog understands the detrimental effect scattering can have on your ability to see. When light interacts with a scattering center, in this case a fog droplet, it is scattered into a new direction, ultimately turning the world around you into a dull gray haze. In some fogs, visibility can be less than 100 meters. It would be possible to see through turbid media like fog if you can separate the scattered light from the unscattered, or ballistic, light; however, we must understand the light transport properties of the atmosphere to determine the optimum scheme. Here, we present an end-to-end simulation for polarized light transport through fog. Our approach can be summarized in three steps: compute the Mueller matrix for a single scattering interaction, ensemble average a distribution of sizes and shapes, and solve the light transport using a Monte Carlo simulation. For small spherical particles, such as fog, we use Mie theory to calculate the single scattering Mueller matrix, but this approach can be generalized to non-spherical particles using ray tracing for large particles or a T-matrix approach for smaller particles. Through this simulation, we are able to determine a backscattering Mueller matrix and a forward scattering Mueller matrix response function for the atmosphere as a function of position and detection angle.
We demonstrate both theoretically and experimentally that a proper surface modification of a highly scattering material can substantially improve light coupling into turbid medium and increase the photon life-time in this medium. As a practical example, we demonstrate that such amended excitation geometry leads to a factor of 100 improved Raman signal efficiency.
Light propagation in a turbid medium is typically considered for flat or regular surfaces. However, such an approximation often does not reflect an experimental reality, and, in this report, we attempt to optimize the surface of a scattering medium to improve the optical coupling into the medium. By making conical microchannels in a turbid medium using short-pulsed laser micro-drilling, we show that we were able to substantially increase the photon life-time and diffusion radius in the medium.
Noninvasive glucose sensing is a Holy Grail of diabetes mellitus management. Unfortunately, despite a number of innovative concepts and a long history of continuous instrumental improvements, the problem remains largely unsolved. Here we propose and experimentally demonstrate the first successful implementation of a novel strategy based on vibrational overtone circular dichroism absorption measurements. Such an approach uses a short-wavelength infrared excitation (1000-2000 nm), which takes the advantage of lower light scattering and intrinsic chemical contrast provided by the chemical structure of D-glucose molecule. We model the propagation of circular polarized light in scattering medium using Monte Carlo simulations to show the feasibility of such approach in turbid medium and demonstrate the proof of principle using optical detection. We also investigate the possibility of using ultrasound detection through circular dichroism absorption measurements to achieve simple and sensitive glucose monitoring.
Second-harmonic generation (SHG) has proven to be an effective method to both image and detect structural
variations in fibrillar collagen. The ability to detect these differences is especially useful in studying diseases
like cancer and fibrosis.1 SHG techniques have historically been limited by their ability to penetrate and image
through strongly scattering tissues. Recently, optical wavefront shaping has enabled light to be focused through
highly scattering media such as biological tissue.2-4 This technology also enables us to examine the dependence
of second harmonic generation on the spatial phase of the pump laser. Here, we demonstrate that wavefront
shaping can be used to enhance the generation of second harmonic light from collagen fibrils even when scattering
is low or non-existent.
The ability to non-invasively focus light through scattering media has significant applications in many fields ranging from nanotechnology to deep tissue sensing. Until recently, the multiple light scattering events that occur in complex media such as biological tissue have inhibited the focusing ability and penetration depth of optical tools. Through the use of optical wavefront shaping, the spatial distortions due to these scattering events can be corrected, and the incident light can be focused through the scattering medium. Here, we demonstrate that wavefront shaping can be used to non-invasively enhance the Raman signal of a material through a scattering medium. Raman signal enhancement was achieved using backscattered light and a continuous sequential algorithm. Our results show the potential of wavefront shaping as an important addition to non-invasive detection techniques.
Wide-field microscopy, where full images are obtained simultaneously, is limited by the power available from speckle-free light sources. Currently, the vast majority of wide-field microscopes use either mercury arc lamps, or LEDs as the illumination source. The power available from these sources limits wide-field fluorescent microscopy to tens of microseconds temporal resolution. Lasers, while capable of producing high power and short pulses, have high spatial coherence. This leads to the formation of laser speckle that makes such sources unsuitable for wide-field imaging applications. Random Raman lasers offer the best of both worlds by producing laser-like intensities, short, nanosecond-scale, pulses, and low spatial coherence, speckle-free, output. These qualities combine to make random Raman lasers 4 orders of magnitude brighter than traditional wide-field microscopy light sources. Furthermore, the unique properties of random Raman lasers make possible the entirely new possibilities of wide-field fluorescence lifetime imaging or wide-field Raman microscopy. We will introduce the relevant physics that give rise to the unique properties of random Raman lasing, and demonstrate early proof of principle results demonstrating random Raman lasing emission being used as an imaging light source. Finally, we will discuss future directions and elucidate the benefits of using random Raman lasers as a wide-field microscopy light source.
Monte Carlo simulations are widely considered to be the gold standard for studying the propagation of light in turbid media. However, due to the probabilistic nature of these simulations, large numbers of photons are often required in order to generate relevant results. Here, we present methods for reduction in the variance of dose distribution in a computational volume. Dose distribution is computed via tracing of a large number of rays, and tracking the absorption and scattering of the rays within discrete voxels that comprise the volume. Variance reduction is shown here using quasi-random sampling, interaction forcing for weakly scattering media, and dose smoothing via bi-lateral filtering. These methods, along with the corresponding performance enhancements are detailed here.
Monte Carlo simulations are widely considered to be the gold standard for studying the propagation of light in turbid media. However, traditional Monte Carlo methods fail to account for diffraction because they treat light as a particle. This results in converging beams focusing to a point instead of a diffraction limited spot, greatly effecting the accuracy of Monte Carlo simulations near the focal plane. Here, we present a technique capable of simulating a focusing beam in accordance to the rules of Gaussian optics, resulting in a diffraction limited focal spot. This technique can be easily implemented into any traditional Monte Carlo simulation allowing existing models to be converted to include accurate focusing geometries with minimal effort. We will present results for a focusing beam in a layered tissue model, demonstrating that for different scenarios the region of highest intensity, thus the greatest heating, can change from the surface to the focus. The ability to simulate accurate focusing geometries will greatly enhance the usefulness of Monte Carlo for countless applications, including studying laser tissue interactions in medical applications and light propagation through turbid media.
Anderson localization, also known as strong localization, is the absence of diffusion in turbid media resulting from wave interference. The effect was originally predicted for electron motion, and is widely known to exist in systems of less than 3 dimensions. However, Anderson localization of optical photons in 3 dimensional systems remains an elusive and controversial topic. Random Raman lasing offers the unique combination of large gain and virtually zero absorption. The lack of absorption makes long path length, localized modes preferred. The presence of gain offsets what little absorption is present, and preferentially amplifies localized modes due to their large Q factors compared with typical low Q modes present in complex media. Random Raman lasers exhibit several experimentally measured properties that diverge from classical, particle-like, diffusion. First, the temporal width of the emission being 1 to a few nanoseconds in duration when it is pumped with a 50 ps laser is a full order of magnitude longer than is predicted by Monte Carlo simulations. Second, the random Raman laser emission is highly multi-mode, consisting of hundreds of simultaneous lasing modes. This is in contrast to early theoretical results and back of the envelope arguments that both suggest that only a few modes should be present. We will present the evidence that suggests a divergence from classical diffusion theory. One likely explanation, that is consistent with all of these anomalies, is the presence of high-Q localized modes consistent with Anderson localization.
The time-temperature effects of laser radiation exposure are investigated as a function of wavelength. We experimentally measure the thermal response of tissue to laser radiation ranging in wavelength from 1100 nm to 1550 nm. Simulations were then performed to estimate damage thresholds.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.