This PDF file contains the front matter associated with SPIE Proceedings Volume 10873 including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

We present the development of a cost-effective, sensitive and specific diagnostic methodology to improve the reliability of the cytopathology diagnosis. The methodology will be based on a new cytology protocol where immunolabeling is performed on fresh cells before fixation using spectrally distinctive plasmonic NPs conjugated with antibodies as optical biomarkers. Metallic NPs, typically gold, silver and Au/Ag alloys, are widely used due to their unique plasmonic properties, photo-stability, water solubility and biocompatibility for in vitro and in vivo biomedical applications. The very distinctive NPs chromatic signature depends on their composition, size, and geometry and provides excellent opportunities for a reliable multicolor imaging and multiplexed immunolabeling. The presented methodology includes a new multispectral and hyperspectral dark-field microscopy for reliable multiplexed and quantitative immunoplasmonic markers optical detection. We applied two optical encoding strategies of immunoplasmonic microscopy (IPM) for immunoplasmonic NPs detection in the NPs-cells complex. The first method is based on reflected light microscopy mode (Patskovsky, S. et al J Biophotonics 8 (5), 401-407 (2015))combined with compact hyperspectral scanning source. It provides spectral differentiation, precise spatial localization and multiplexed quantification of NPs labels. The second approach uses a multispectral side-illumination dark-field microscopy that allows to design a compact module for optical imaging and spectroscopic identification of individual plasmonic NPs in fixed or live cell preparations. The presented approach can provide a convenient and routine method for immunoplasmonic markers visualization by the pathologist. It can be easily adaptable to the microscopes currently used in the clinical setting thus facilitating and accelerating its adoption.

Diabetic Foot Ulcers (DFUs) are responsible for 20% of diabetic-related hospitalization and 85% of diabetes related amputations. In DFUs the primary factor affecting healing is an adequate oxygen supply to the wound. However, the gold standard approach for assessing DFUs is by evaluating the reduction of wound size over a four-week period. In this study, we investigate the potential of altered breathing patterns as a technique to assess localized oxygenated perfusion in DFUs as a measure of healing potential. A continuous wave (CW), non-contact, near infrared optical scanner (NIROS) was used to conduct NIR based spectroscopic imaging at dual discrete wavelengths (729nm and 799nm) on DFUs with 7mW of maximum optical power. Subjects were imaged at discrete time points and dynamically utilizing an altered breathing paradigm (i.e. breath-hold) to measure the relative oxy- (ΔHbO) and deoxyhemoglobin (ΔHbR) changes in normal and DFU scenarios. Results show that in normal individuals, ΔHbO/ΔHbR changes at all points of the foot because of altered breathing patterns are synchronous; whereas in the DFU scenario changes in hemodynamic parameters are asynchronous. This indicates that under normal circumstances, oxygenated perfusion changes are consistent and uniform at all points of the foot as opposed to the DFU scenario’s inconsistent oxygenated perfusion. Altered breathing paradigms may serve as a useful tool in assessing localized sub-surface oxygenated perfusion in regions around the wound, and help clinicians better cater the treatment process.

Near-infrared (NIR) spectroscopic imaging of wounds has been performed by past researchers to obtain tissue oxygenation at discrete point locations. We had developed a near-infrared optical scanner (NIROS) that performs noncontact NIR spectroscopic (NIRS) imaging to provide 2D tissue oxygenation maps of the entire wounds. Regions of changed oxygenation have to be demarcated and registered with respect to visual white light images of the wound. Herein, a semi-automatic image segmentation and co-registration approach using machine learning has been developed to differentiate regions of changed tissue oxygenation. A registration technique was applied using a transformation matrix approach using specific markers across the white light image and the NIR images (or tissue oxygenation maps). This allowed for physiological changes observed from hemodynamic changes to be observed in the RGB white light image as well. Semi-automated segmentation techniques employing graph cuts algorithms was implemented to demarcate the 2D tissue oxygenation maps depicting regions of increased or decreased oxygenation and further coregistered onto the white light images. The developed registration technique was validated via phantom studies (both flat and curved phantoms) and in-vivo studies on controls, demonstrating an accuracy >97%. The technique was further implemented on wounds (here, diabetic foot ulcers) across weeks of treatment. Regions of decreased oxygenation were demarcated, and its area estimated and co-registered in comparison to the clinically demarcated wound area. Future work involves the development of automated machine learning approaches of image analysis for clinicians to obtain real-time co-registered clinical and subclinical assessments of the wound.

Experimental design: We used Raman spectroscopic analysis of human prostate cancer patient tissues to characterize composition of lipid droplets in the metastatic lesions. The therapeutic efficacy was tested in prostate cancer orthotopic and intra-cardiac injection mouse models. Gene expression profiling was used to identify genes that are affected by cholesteryl ester depletion. Additionally, immunoblotting, immunofluorescent staining and transwell assay were used to further verify inactivation of Wnt/β-catenin pathway by cholesteryl ester depletion. Stimulated Raman scattering microscopy and mass spectrometry were used to assess lipogenic potential of cholesteryl ester-depleted prostate cancer cells. Results: We observed accumulation of cholesteryl ester in metastatic lesions of human prostate cancer patient tissues. Inhibiting cholesterol esterification significantly reduced the number of metastatic clusters by 50% in the prostate cancer orthotopic mouse model. We showed suppression of metastatic tumor growth in the intra-cardiac injection model without observable toxicity to the mice. The mechanism study further supports that cholesteryl ester depletion suppresses metastasis by upregulation of regulators that negatively impact prostate cancer metastasis. Notably, Wnt/β-catenin is downregulated upon cholesteryl ester depletion, and we found evidence suggesting that cholesteryl ester depletion significantly blocks secretion of Wnt3a through reduction of monounsaturated fatty acid levels, which limits Wnt3a acylation. In conclusion, we demonstrate that targeting cholesterol esterification can treat metastatic prostate cancer effectively with minimum toxicity.

Optical spectroscopic techniques based on tissues, cell lines and biofluids for the detection of cancer and various disorders are being extensively opted due to their high sensitivity, ease of procedure and sample preparation. Though the use of fluorescence in the field of oncology started as early as 1924 by Policard et al, the use of Native Fluorescence Spectroscopy (NFS) as a tool for cancer diagnosis and treatment monitoring, has shown significant progress only in the recent years, due to the advancements in the detection system and software. Even though an array of chemotherapeutic agents and sophisticated Radiotherapy techniques have evolved in the management of cancers, surgery is mostly opted as the first line of treatment in most of the breast cancer cases, in an approach to remove the bulk of tumour from the body. The goal of this study is to assess the efficacy of various treatment modalities, in this case surgery (modified radical mastectomy) and chemotherapy by comparing the NF spectra of breast cancer patients, post-surgery with those on completion of chemotherapy and normal subjects. In this pilot work, NF spectroscopy of blood plasma of clinically confirmed 43 breast cancer patients were measured under pre-treatment, post-surgical, post chemotherapy and also during yearly follow-up (after completion of the entire treatment) conditions. The NFS spectra of 60 normal subjects were also recorded for comparison. The results indicate that there is a statistical significance in the spectral variations in the region of collagen, elastin and NADH between the normal and cancer patient’s blood plasma as well as between pre and post treated cases as well as those patients on yearly follow-up after completion of the entire treatment.

Understanding the dynamics of metabolism in a multicellular organism is essential to unraveling the mechanistic basis of many biological processes. It is the synthesis, transformation and degradation of biomolecules (the definition of metabolism) that carry out the genetic blueprint, which cannot be imaged in vivo by using traditional methods. In the present work, we developed a new method that combines D2O probing and Stimulated Raman Scattering microscopy (DO-SRS) to visualize metabolic dynamics in live animals. The enzymatic incorporation of D2O-derived deuterium (D) into biomolecules will generate carbon-deuterium (C-D) bonds in macromolecules. Within the broad vibrational spectra of C-D bonds, we discover lipid-, protein-, and DNA-specific Raman shifts and develop spectral unmixing methods to obtain C-D signals with macromolecular selectivity. We obtained new biological insights in several studies such as the spatial dependence of lipogenic activities of sebaceous glands, specific myelination timing of the developing mouse brain, differential protein and lipid metabolism in germline development of C. elegans as well as its aging process, the spatial constrain for the distribution of newly synthesized yolk proteins in aged C. elegans, the prevalence of protein biosynthesis and the lack of lipogenesis in zebrafish embryos, and intratumoral metabolic heterogeneity. In summary, we demonstrated that our current DO-SRS method is better than other deuterium-labeled carbon substrate in monitoring and imaging metabolic activities. This technique can track specifically de novo lipogenesis, image in vivo protein biosynthesis without tissue bias, and can simultaneously image spatial temporal dynamics of lipid and protein.

The hemodynamics and oxygen saturation status of vascular are very important biomarkers for disease, such as brain glioma tumor and ischemia-reperfusion ulcer. Therefore, a high spatial resolution imaging tool for vascular imaging is demanded. Conventional optical imaging modalities, including confocal microscopy and two-photon microscopy, require external contrast agent to image blood vessels and are not sensitive to oxygen saturation. The development of photoacoustic microscopy provides a contrast-free, high-spatial resolution and functional vascular imaging tool. It’s gaining more and more popularity in biomedical research. In this paper, we introduce a dual-wavelength opticalresolution photoacoustic microscopy (OR-PAM) system for functional imaging of vasculature. This system has demonstrated its application in brain glioma tumor imaging, as well as skin ischemia-reperfusion imaging.

We demonstrate optical biopsy of vocal folds to assess its surface and subsurface movement during phonation on ex-vivo calf larynx using parallel OCT (POCT). In this technique, we will enable simultaneous interrogation of multiple locations along the vocal fold, there by eliminating motion-blur artifacts exhibited by the sequential sampling provided by conventional flying-spot OCT. Currently, in voice clinics, laryngologists and speech-language pathologists rely heavily on Video Stroboscopy (VS) coupling with, visual judgments of vocal fold morphology and auditory perceptions of voice quality, to make effective diagnostic, surgical and therapeutic decisions. There is a constant need of an endoscopic imaging tool in voice clinics that can directly capture the three-dimensional (3D) surface motion of the vocal folds in real time as patients phonate. To address the need, we will present a POCT/VS imager that will combine parallel swept source OCT technique with VS to provide a real time display of the vocal folds in all three axial dimensions during phonation. The results will yield cross sections (B-Scans) with ~16 co-linear sampling locations spread over ~5mm on a phonating ex-vivo calf larynx showing fluid periodic cyclic motion of the vocal folds (~A-scan rate) in real time enabled by POCT approach. We will also capture VS images in sync with POCT B-scans validating the real time cross sectional probing of POCT/VS imager.

In this paper, we summarize our recent advances in the development and pre-clinical testing of the 2-nd generation OCT-based probe for core needle biopsy guidance. The acquired OCT images are processed in real-time in a GPU unit to provide the interventional radiologist with the capability to examine the tissue cellularity at the tip of the biopsy needle before deciding to take a biopsy core. The extensive testing of this technology on a rabbit model of soft tissue cancer is discussed in detail. The pre-clinical results demonstrate the capability of this OCT-based probe for determining the in situ cellularity of the tissue at the tip of the biopsy needle and thus its potential use for improving the quality of the sampled biopsy cores.

Mueller Polarimetric Imaging (MPI) showed promising results in biomedical applications, especially for early detection of precancerous lesions on biological tissues. This technique is label-free, non-invasive and can be implemented with a large field of view (up to several cm²) to image wide areas of biological tissues while providing information on its microstructure. The development of innovative (MPI) systems, able to analyze biological tissues in vivo on human patients, remains an instrumental challenge. Our goal is to build miniaturized and compact full-field MPI systems based on Ferroelectric Liquid Crystals (FLCs) capable of performing a multispectral accurate analysis of biological tissues in vivo. In this work, an innovative approach is showed to realize optimized and fast FLCs-based MPI systems able to perform full-field imaging acquisitions in the spectral range between 450 and 700nm with error less than 1% on all the elements of measured Mueller matrices. This system can be accurately calibrated by using the Eigenvalue Calibration Method (ECM) also in presence of high residual instrumental depolarization. This approach enables us to realize compact and reliable MPI systems which can be easily integrated into existing instruments currently used in medical practice.

In turbid tissue-like scattering medium the conventional polarized light, scattered multiple number of times, is depolarized, and the depolarization rate depends strongly on the size and shape of scattering particles, as well as on the number of scattering events. In fact, the structure of light can be more complicated when the polarization of light across the laser beam can be radially or azimuthally polarized and carry orbital angular momentum. When these vector light beams, known as cylindrical vector beam (CVB) and Laguerre-Gaussian (LG) beams, propagates in turbid tissue-like scattering medium, either anisotropic or inhomogeneous, the spin or angular momentum are changed that leads to spin-orbit interaction. Such a spin-orbit interaction leads to the mutual influence of the polarization and the trajectory of the light propagation. We investigate the applicability of using CVB and LG beams for optical biopsy. In current presentation propagation of CVB and LG beams in anisotropic turbid tissue-like scattering media is considered in comparison to conventional Gaussian beams. We demonstrate that by applying CVB and LG beams the sensitivity of the conventional polarimetry-based approach is increased at least twice in comparison with the experiments utilizing ‘simple’ Gaussian polarized light. The results of the study suggest that there is a high potential in application of vector light beams in tissue diagnosis.

Biological tissues characterization can be approached by non-ionizing optical techniques, in a non-invasive, non-contact way. Optical diagnostic techniques include Optical Coherence Tomography, spectroscopy or fluorescence, among others. Tissue differentiation is difficult to achieve in general with high specificity and sensibility. Spectroscopy is of great interest for this aim, as it provides intrinsic molecular contrast. The different composition and/or structure of biological tissues influence the spectral response. However, the interpretation of spectra is difficult from the raw data, and further data processing is needed. Diffuse Reflectance Spectroscopy (DRS) is particularly well-suited for biomedical applications, as it can work with bulk tissues in reflection, reinforcing the non-invasive character of the technique. DRS has been employed for malignant tissue detection and also for healthy tissue discrimination. These applications require an adequate definition of potential biomarkers for the classification algorithms. The classification process depends strongly on the amount of collected spectra and tissue and specimen variability. In this work several types of ex-vivo porcine tissues are extracted and measured by DRS. Spectral measurements are made on different specimens, and on different points of each sample. Spectra are normalized and several algorithms for dimension and variability reduction are applied, such as Principal Component Analysis or Savitzky-Golay filtering. From these spectra, several biomarkers are proposed for tissue classification, and different classifiers are applied. The results are compared, and the classification efficiency is quantified. The considered approaches could be of particular interest in image-guided surgery or other types of optical biopsy applications.

The spectral and spatial resolution of hyperspectral imaging is useful for investigation of tissue autofluorescence. The low-light, noisy conditions in fluorescence imaging usually necessitates noise removal for extraction of precise spectral signatures and peak shifts. However, noise removal techniques like low-pass filtering or the Maximum Noise Fraction transform might discard information or distort spectral features. In this study, smoothing splines is proposed as an alternative technique to avoid spectral distortion in analysis of hyperspectral fluorescence images in the wavelength range 400-1000 nm. Continuous tuning parameters and use of natural cubic splines makes the method advantageous for unbiased peak extraction. The method was tested on ex vivo images of atherosclerosis lesions and simulations. The method was used to estimate autofluorescence peak shifts, and found to perform well in comparison with MNF.

Fluorescence lifetime imaging microscopy (FLIM) techniques are widely used in auto-fluorescence imaging to investigate dynamic metabolic states of the cells. Traditional FLIM imaging requires the cells to adhere to the coverslip to collect enough photons for FLIM data processing. Such conditions pose challenges to non-adherent cells, such as acute myeloid leukemia (AML) cells, because of cell motility. We developed a proto-type micro cell trapping array (MCTA) to immobilize cells in picoliter-size wells with location references. The array keeps non-adherent cells in referenced well locations, allows treatment on stage and re-imaging after time for ultimate cell segmentation analysis. Individual wells are analyzed by a pixel-based region-of-interest (ROI) to analyze cellular redox states. This single well trapping and analysis method allows to isolate treatment responses of a small number of cells, compare their range and predict early effect, which may have clinical applications in the context of cancer aggressiveness and treatment outcomes. We expanded the common intensity-based assay for cellular redox state by a Fluorescence Lifetime measurement, NAD(P)H-a2%, serving as an alternative metric. The new assay is flexible and can be applied to other non-adherent cell lines, expanding FLIM applications in both research and the clinic.

Restoring normal functioning and tissue healing after surgical intervention is, among others, critically dependent on tissue oxygenation and perfusion. Tissue necrosis, caused by inadequate tissue perfusion and/or oxygenation, is a common complication after surgical reconstruction of e.g. bowel anastomoses or skin defects. Currently, several optical techniques, which do not require administration of contrast agents, have been used to evaluate tissue perfusion and oxygenation. An emerging technique is hyperspectral imaging, which is capable to detect the scattering and absorption of light delivered to the tissue, caused by inhomogeneity of biological structures, such as haemoglobin, fat or water. Recently, a snapshot hyperspectral camera was developed to measure tissue oxygenation non-invasively using relevant wavelengths in the VIS-NIR region (450-950 nm). In this study, the effect of occlusion-reperfusion of the brachial artery on cutaneous blood oxygenation is explored in three human volunteers as assessed by a snapshot hyperspectral camera system. Furthermore, measurements of local changes in skin oxygenation and blood flow after applying a local vasodilator (capsaicin-based cream) and a local vasoconstrictor (brimonidine gel) are compared to measurements of an untreated area of the skin. Hyperspectral results are correlated to the haemoglobin oxygen saturation measured by an oximeter. Simultaneously with the hyperspectral measurements, real-time blood perfusion mapping is performed using Laser Speckle Contrast Imaging, which is able to measure cutaneous skin blood flow through analysis of the speckle pattern.

Time-resolved separation of Raman scattering from background fluorescence is demonstrated using a recently developed 512 pixel, 16.5 giga-events CMOS SPAD line sensor ¹ . The system is being developed with the aim of evaluating the suitability of liver tissue for transplant surgery, as these samples pose particular challenges to commercial Raman systems due to their high fluorescence emission across a wide spectral excitation range. Compared to previous work ² , the spectrometer operates without the use of time-gating, relying entirely on inpixel time-correlated single photon counting (TCSPC). Indeed, by employing the unique features of the sensor, such as on-chip histogramming and zoomable time resolution from 50ps to 6.4ns, the system is able to deliver both Raman and time-resolved fluorescence decay data. Time-resolved separation of Raman and fluorescence signals allows the spectrometer to be operated in the visible range (using a 532-nm pulsed laser), thus providing enhanced Raman scattering intensity compared with the use of a near-infrared laser, since scattering emission is proportional to λ-^{4 3} . The system is calibrated using a Neon calibration source and benchmarked using samples of pure distilled-water fluorescein, paracetamol and sesame oil in comparison with results from continuous wave excitation in a Renishaw InVia spectrometer. The Raman band of water at a Raman shift of 3000-3800 cm-1 is chosen to evaluate sensor performance because of its low intensity and its characteristic spectral profile which is readily compared with the literature ⁴ .

We are introducing an application of a machine learning approach for express analysis of Laser Speckle (LS) images. This application can be utilized for real-time visualisation of vascular beds in vivo. This research used Waikato Environment for Knowledge Analysis (Weka) integrated with Fiji/ImageJ software. A large number of acquired LS images are averaged, then used as references for training Weka classifiers. Subsequently, a bundle of these Weka classifiers are produced. We defined the minimal number of raw LS images based on a phenomenological model to minimize the time needed for LS data analysis. Finally, a new perceptually uniform color coding approach is developed for highlighting targeted blood vessels. The developed LS processing approach is especially convenient, because of its high potential for blood vessel visualisation during real-time intraoperative vascular imaging in vivo.

Most recently, the second (1100-1350 nm), third (1600-1870) and fourth (2100-2350 nm) NIR windows have received increased attention for use in linear and nonlinear imaging and spectroscopy. Due to minimal scattering and absorption at these windows, there is a reduction in blurring and image quality is enhanced. SWIR wavelengths can penetrate deeper than when using conventional methods and are being utilized for medical applications including brain imaging, identifying bone micro-fractures and delineating breast cancer margins. The supercontinuum laser can provide a powerful broadband light source that will increase SWIR applications. Recent advances will be discussed.

Modern traumatic injuries, as encountered in battlefield conflicts, are often characterised by extensive soft tissue damage from blasts and high energy projectiles. This situation has created a challenge for wound stabilisation and repair, with surgical intervention common, via wound debridement procedures. These are often complex surgeries where necrotic and infected tissue is removed, usually with multiple remedial surgeries, designed to aid the natural healing process and to reduce the likelihood of infection. With extensive injuries, the preservation of viable tissue is paramount to functional recovery. Additionally, identifying wounds which are likely to heal without intervention, as well as those that exhibit precursors for impaired healing or infection, would assist in informing the appropriate medical care. Technologies that utilise concepts of non-contact imaging, such as optical imaging and spectroscopy can be used to obtain spatial and spectral maps of biomarkers, which provide valuable information on the wound (e.g. precursors to improper healing or delineate viable and necrotic tissue). A negative contrast imaging device (NCI) has been shown to characterise wound biopsies, through mid-IR (2.6 – 4 μm) non-invasive spectroscopic imaging. To better demonstrate the applicability of this technique, wound relevant cell cultures, subjected to induced trauma, are used to identify spectral changes between healthy and traumatised cells. This work highlights the available contrast in spectroscopic mid-IR signals and demonstrates the utility of spatially and spectrally derived maps as an assessment tool for wound diagnostics.

Optical spectroscopy is a powerful technique that allows for label-free, noninvasive, and real time characterization of biomolecules. Compared with other optical techniques that relies on the shift of a single resonance, such as surface plasmon resonance (SPR) sensors and optical-grating-based cell assays, spectroscopic techniques can discriminate between different chemical species and are suited for analyzing often complex biological samples. Surface-enhanced infrared absorption (SEIRA) based on top-down fabricated substrates such as nanoantennas, nanoslits, and metasurfaces has been demonstrated as a versatile technique that can enhance the IR signal and characterize small amounts of adsorbed protein and lipid films. Here, we demonstrate the use metasurface-enhanced infrared reflection spectroscopy (MEIRS) to observe live cells cultured on top of the plasmonic metasurface. MEIRS has a penetration depth on the order of tens of nanometers, longer than surface-enhanced Raman spectroscopy (SERS) and yet shorter than attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy. This makes MEIRS uniquely suited for probing the surface of a cell, to observe features such as protein expression in the cell membrane and cellular adhesion. This has important implications in the detection of cancer cells through spectroscopic cytology, as cancer cells differ from regular cells significantly in the expression of membrane proteins and adhesion molecules. In this work, we demonstrate the observation of cancer cell adhesion through IR spectroscopic mapping. Furthermore, we show the effect of different anticancer cocktails (doxorubicin, salinomycin, and their combination) on cancer cells, as observed by MEIRS.

Optogenetic stimulation of neurons, driven by the development of light-gated rhodopsins, has revolutionized the experimental interrogation of neural circuits and holds promise for next-generation treatment of neurological disorders. However, it is limited by the inability of visible light to penetrate deep inside brain tissue. Optical stimulation of deep brain neurons, for example, has hitherto required the insertion of invasive optical fibers because the activating blue-green wavelengths are strongly scattered and absorbed by endogenous chromophores. Red-shifted variants of rhodopsins have been developed, but their action spectra still fall out of the near-infrared (NIR) optical window (650-1350 nm) where light has its maximal depth of penetration in brain tissue. Here, we developed a novel approach for NIR optogenetics, where lanthanide-doped upconversion nanocrystals (UCNPs) were used to absorb tissue-penetrating 980 nm NIR and emit visible light for rhodopsin activation. Due to lanthanides’ ladder-like electronic energy structure, the emission of UCNPs can be precisely tuned to a particular wavelength by control of energy transfer via selective lanthanide-ion doping. For instance, incorporation of Tm³⁺ into Yb³⁺ doped host lattices leads to blue emission (~470 nm) that matches the maximum absorption of channelrhodopsin-2 (ChR2) for neuronal activation, while the Yb³⁺/Er³⁺ couple emits green light (~540 nm) compatible with activation of halorhodopsin (NpHR) or archaerhodopsin (Arch) for neuronal inhibition. We demonstrated that molecularly tailored UCNPs could serve as optogenetic actuators of transcranial NIR to functionally stimulate deep brain neurons in mice. Transcranial NIR UCNP-mediated optogenetics evoked dopamine release from genetically tagged neurons in the ventral tegmental area, induced brain oscillations via activation of inhibitory neurons in the medial septum, silenced seizure via inhibition of excitatory cells in the hippocampus, and triggered memory recall via excitation of a hippocampal engram. UCNP technology would open the door to less-invasive optical neuronal activity manipulation with the potential for remote therapy.

We present a mid-infrared spectral-domain optical coherence tomography system operating at 4.1 μm central wavelength with a high axial resolution of 8.6 μm enabled by more than 1 μm bandwidth from 3.58-4.63 μm produced by a mid-infrared supercontinuum laser. The system produces 2D cross-sectional images in real-time enabled the high-brightness of the supercontinuum source in combination with broadband upconversion of the signal to the range 820-865 nm, where a standard 800 nm array spectrometer can be used for fast detection. We discuss the potential applications within nondestructive testing in highly scattering materials and within biomedical imaging for achieving the in-vivo optical biopsy.

Hyperspectral imaging is a generic imaging modality allowing high spectral and spatial resolution over a wide wavelength range from the visible to mid-infrared. Short wavelength infrared (SWIR) hyperspectral imaging is currently becoming an important supplement to spectroscopy in optical diagnostics due to the flexibility and adaptability of the technique. However, due to the complexity of hyperspectral data, the analysis requires a well planned approach. In this paper a simple but effective approach combining dimension reduction and unsupervised classification is suggested. Examples of in vivo hyperspectral data in the SWIR spectral range (950-2500 nm) from human skin bruises and porcine skin burns are presented as examples. Data are processed using the minimum noise fraction transform (MNF), and K-means clustering. K-means clustering was found to perform significantly better if applied to MNF transformed data. The classification results agree well with biopsies, spectral data and visual inspection of injuries. It is thus shown that unsupervised clustering can be a preferable technique in cases where it is challenging to use or interpret results from physics based models, or where the ground truth is lacking or not well defined. The presented results confirm that SWIR hyperspectral imaging indeed is a useful tool for optical characterization of tissue.

The development of new, compact mid-infrared light sources is critical to enable biomedical sensing applications in resource-limited environments. Here, we review progress in fiber-based mid-IR sources, which are ideally suited for clinical environments due to their compact size and waveguide format. We first discuss recent developments in mid-IR supercontinuum sources, which exploit nonlinear optic phenomena in highly nonlinear materials (pumped by ultrashort pulse lasers) to generate broadband spectra. An emerging alternative approach is then presented, based on broadly tunable mid-IR fiber lasers, using the promising dysprosium ion to achieve orders of magnitude higher spectral power density than typical supercontinua. By employing an acousto-optic tunable filter for wavelength tuning, an electronically controlled swept-wavelength mid-IR fiber laser is developed, which is applied for absorption spectroscopy of ammonia (NH₃), an important biomarker, with 0.3 nm resolution and 40 ms acquisition time.

A new paradigm in mid-infrared (MIR) biophotonics is opening up. The idea is of portable, MIR hyperspectral imaging in real-time for healthcare, including for in vivo cancer screening and diagnosis, based on new MIR fiber-optics. At the heart of the MIR fiber-optic approach are fiber-based MIR-supercontinuum (SC) broadband laser sources. First time reports of tissue hyperspectral imaging using fiber-based MIR-SC laser sources are reviewed here.

Convolutional neural networks (CNN) are a class of machine learning model that are especially well suited for imagebased tasks. In this study, we design and train a CNN on tissue samples imaged using Multi-Photon Microscopy (MPM) and show that the model can distinguish between chromophobe renal cell carcinoma (chRCC) and oncocytoma. We demonstrate the method to train a model using simple max-pooling vote fusion, and use the model to highlight regions of the input that cause a positive classification. The model can be tuned for higher sensitivity at the cost of specificity with a constant threshold and little impact to accuracy overall. Several numerical experiments were run to measure the model’s accuracy on both image and patient level analysis. Our models were designed with a dropout parameter that biases the model towards higher sensitivity or specificity. Our best performance model, as measured by area under the receiver operating characteristic curve (AUC of ROC, or AUROC) on patient level classification, is measured with a 94% AUROC and 88% accuracy, along with 100% sensitivity and 75% specificity.

We are currently investigating a non-invasive technique to determine collagen content and hydration in the skin by the optical methods of absorption, fluorescence and Raman spectroscopy for imaging. A major barrier to the effective use of skin therapies is the difficulty of quantifying existing collagen content and water content. Absorption of near infrared light by skin depends both on the concentration of collagen and the amount of water in the skin. In the near infrared (NIR window I) collagen and water have similar absorption profiles. However, because the infrared spectrum of collagen and water from 900 nm to 1700 nm (window II and III) are significantly different, it allows us to quantify collagen relative to water content. The ratio of the absorption of collagen normalized to water at 1700 nm and at 1950 nm (window III and window IV) is linear in collagen concentration. This can be used to discriminate between tissues by absorption imaging. We compare these results to Raman spectroscopy and native fluorescence. Our goal is to generate data that can be used for qualitative imaging allowing for improvement in assessing the effectiveness of skin-treatment therapies for the health care field to develop a device for home and medical office which can answer the age-old question: “Mirror, Mirror on the wall, where exactly should I apply skin therapeutics for maximum effectiveness and minimal side-effects?”

In gastrointestinal endoscopic surgery, bleeding from the accidental resection of hidden vessels is a major complication, requiring immediate conversion to open surgery. Methods of visualizing occult vessels have been proposed using the FDA approved fluorescent dye Indocyanine green (ICG), but Native label-free fluorescence of the submucosa—present up till about 886 nm—prevents the use ICG in near-infrared (NIR) window I (700 nm to 900 nm). Instead, absorption imaging is preferred; the darker vessels are visible to 4.5 mm deep. Using data from Raman scattering, absorption, native fluorescence, SHG and the photon excitation fluorescence we investigate the spectral properties and propose optimal parameters for differentiation of blood vessels from surrounding tissue in a variety of tissue types in NIR window II (1000 nm to 1350 nm) and NIR window III (1550 nm to 1900 nm, the “Golden Window”) as a complement to absorption imaging.

Laser-induced fluorescence (LIF) technique was used to generate spectral signatures of endogenous fluorophores relevant to the tissue molecular composition changes in human brain glioma tumors. The goal is to study the changes of fluorescence emission spectra from endogenous fluorophores in human brain glioma of different grades, and to find new biomarkers for prognostic optical molecular pathological diagnosis. Two hundred and thirty-seven (237) native fluorescence spectra from 61 subjects were measured using LabRAM HR Evolution micro photoluminescence (PL) system for four grades of glioma tumors in ex-vivo. The differences of four grades of glioma tumors were identified by the characteristic fluorophores fingerprints under the excitation laser wavelength at UV 325nm. To our best knowledge, this is the first report for human brain study using this technique. The fluorescence peaks of biomarkers with major contribution were found, including tryptophan, collagen, elastin, reduced nicotinamide adenine dinucleotide (NADH), flavin adenine dinucleotide (FAD) and phospholipids that play important roles in the cellular energy metabolism and glycolysis pathway. The ratios of peak intensities and the peak positions in fluorescence spectra of may be used to diagnose human brain diseases or to guide biopsy during surgical resection.

Abnormal tryptophan metabolism is a major factor in Alzheimer’s disease. Thus, understanding the role of indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO), enzymes that causes increased tryptophan metabolism down the kynurenine pathway, is crucial. Tissue samples from the hippocampus, Brodmann’s area 9 (BA 9) and Brodmann’s area 17 (BA 17) from Alzheimer’s patients, from patients with psychosis and from age-matched, normal controls (total of 47 samples) were studied. An update on the optical properties of Alzheimer’s tissues, utilizing tryptophan/kynurenine ratios, is reported. Ratios of tryptophan/kynurenine were decreased in the hippocampus and BA 9 in patients with Alzheimer’s, consistent with increased neuroinflammation and IDO and/or TDO activity in these areas.

The translation of microscopy with ultraviolet surface excitation (MUSE) into a high school science classroom is investigated with the goal of providing a suitable new modality to enhance life science education. A key part of this effort is the development of laboratory exercises that can integrate the advanced capabilities of MUSE into a classroom setting. MUSE utilizes the unique property of ultraviolet light at wavelengths between 250 and 285 nm to propagate about 10 μm into tissues, thus illuminating only the top cell layer. This illumination is provided by a low-power UV LED source, which enables one to cost-efficiently implement this method into the educational environment. MUSE in education can eliminate the need for premade microscope slides and provide a far more engaging and rewarding experience for students.

Intrinsic fluorescence spectra of fresh normal and cancerous human breast tissues were measured using two selective excitation wavelengths including 290nm and 340nm. Dual-wavelength excitation may reveal more molecular information than single-wavelength excitation. In the meantime, it is significantly faster than the acquisition of excitation-emission (EEM) matrix. Unsupervised machine learning algorithms principal component analysis (PCA) and non-negative matrix factorization (NMF) were used to reduce the dimensionality of the spectral data. The relative concentrations of the basis spectra retrieved by PCA and NMF were considered features of the samples and used to distinguish normal and malignant tissues. The performances of classification using support vector machine (SVM) based on PCA and NMF features were compared. The classification using spectral data with dual-wavelength excitation was compared with single-wavelength excitation. Classification based on NMF-retrieved components from spectral data with dual-wavelength excitation yielded the best performance.