No abstract

No abstract

We present a new and innovative shortwave infrared (SWIR) hyperspectral imaging focal plane array (FPA) concept for bulk and trace standoff explosives detection. The proposed technology combines conventional uncooled InGaAs based SWIR imaging with the wavelength selectivity of a monolithically integrated solid-state Fabry-Perot interferometer. Each pixel of the array consists of a group of sub-pixels in which each sub-pixel is tuned to absorb a separate wavelength. The relative responses from the sub-pixels (i.e. wavelengths) are compared to the spectral characteristics of explosives in the SWIR to detect and locate them within an imaged scene among various background materials. The novel technology will be compact, and consume low power such that it can be used as a handheld device or mounted for persistent surveillance of crowded areas and checkpoints. The technology does not use any scanning nor tuning apparatuses such as MEMS devices, and is therefore fast, compact, lightweight and not susceptible to vibration. The technology is therefore ideal for man portable applications and unmanned vehicle platforms. An eyesafe (covert) illuminator may be used to provide illumination in situations when ambient light conditions are not sufficient. We will present a detailed design of the novel focal plane array and a theoretical standoff distance and false rates study.

We measured the polarization dependence of light scattered from a tilted fiber grating and found disagreement with previous volume-current perturbation analysis. However, by including the longitudinal E field of the guided wave we were able to obtain good agreement, demonstrating that, although it is small, this component cannot be neglected when scattering of weakly guided waves is considered. A first-order approximation formula for the polarization dependence was also obtained and is shown to be accurate within most of the resonance band of scattering.

Novel optoelectronic instrumentation has been developed for the multispectral imaging of autofluorescence emitted by metabolic fluorophores. The images resolve individual cells while spectra are collected for each pixel in the images. These datacubes are generated at a rate of 10 per second-fast enough for surgical guidance. The data is processed in real time to provide a single color-coded image to the surgeon. To date, the system has been applied to fresh, ex vivo, human surgical specimens and has distinguished breast cancer from benign tissue. The approach is applicable to in vivo measurements of surgical margins and needle-based optical biopsies. Ongoing work demonstrates that the system has great potential for translation to a hand-held probe with high sensitivity and specificity. K E Y W O R D S autofluorescence, breast cancer, multispectral imaging, surgical margin more prevalent than typical aerobic cellular respiration through the electron-transport chain. This in turn leads to an increase in the reduced form of nicotinamide adenine dinucleotide (NADH)/flavin adenine dinucleotide (FAD) ratio (or the redox ratio). Simultaneously, hemoglobin oxygenation near the tumor decreases as the growing J Cell Physiol. 2019;234:5413-5419. wileyonlinelibrary.com/journal/jcp © 2018 Wiley Periodicals, Inc. | 5413 CARVER ET AL. | 5417Once fully translated, our optics will be redesigned into an ergonometric hand-held probe tethered to a cart supporting a colorcoded display for use by surgeons in the operating room. This new capability will shift clinical practice by enabling surgeons to assess tumor margins in real time to inform the extent of surgical resection.Color-coded images will cover a 5 mm field of view in 0.1 s. Translation of this capability to a hand-held probe is crucial for implementation by surgeons (You et al., 2018). Validation studies will be extended beyond correlations with H&E staining. We will also harvest cancer and normal cells identified by our optical system via LCM for subsequent analysis by genetic (DNA-encoded regulatory information) and epigenetic (non-DNA-encoded regulatory information) approaches. These efforts are expected to show that our optical contrast is related to cellular metabolism rather than blood and oxygenation issues. It is anticipated that the multispectral imaging approach we have developed for breast cancer will support optical imaging of other solid tumors.Our results were obtained with spectral vectors that were obtained with a software tool that averages the spectra of all pixels in the given image. As a result, a cancer vector can be "polluted" with noncancerous pixels. Even under that circumstance, the clear differences between the two images in Figure 3 are still observed.Ongoing work will "purify" the vectors by averaging over restricted regions-of-interest rather than the full images. The vectors can also be improved by harvesting more wavelengths in the NADH band, and using fewer, wider bins in the FAD band. These enhancements should allow for the straightforward s...