This PDF file contains the front matter associated with SPIE Proceedings Volume 11251, including the title page, copyright information, table of contents, and author and conference committee lists.

Repair or reconstruction of organs is the goal of regenerative medicine. Bioengineered organoids that can differentiate when implanted in-vivo to partially restore organ function are being developed. Potentially, such organoids can be used to treat many medical conditions. A non-invasive method for quality monitoring of tissue engineered constructs is needed in order to ensure that they are ready for implantation. Raman micro-spectroscopy offers a way to quantitatively analyze cells and tissues without sample preparation or labelling dyes, which are not allowed in constructs used for the human implantation. Epithelial progenitor cells are parts of the complex organoids derived from the embryonic salivary gland cells. We have collected Raman spectra of the epithelial (acinar and ductal) cells treated with Fibroblast Growth Factor 2 (FGF2) and grown in organoids ex vivo over a period of (1 - 7 days). Evolution of the organoids over time was detected with Raman. These modifications, corresponding to the C-C stretch and C-H bend in proteins, as well as alterations in the Amide I and III envelopes, likely may correlate with changes in the cell environment or their differentiation state. Our goal is to develop Raman metrics that can be applied to the non-invasive monitoring of organoids.

Raman microscopy is well recognized as a nondestructive, label-free biomedical imaging method that provides abundant chemical information of the specimen. Excitation wavelengths in deep near-infrared (e.g., 1064 nm) are used in certain situations, such as when analyzing photosensitive/photolabile specimens to suppress the strong fluorescence and to avoid photodamage. However, the speed and quality of 1064 nm Raman imaging suffers from the low scattering efficiency at this long excitation wavelength and the high noise level of InGaAs detectors. In this study, we investigated a multifocal patterned approach for 1064 nm Raman imaging. A 2-D Hadamard-coded multifocal array generated with X-Y scanning galvomirrors is used to excite and collect multiple Raman spectra simultaneously. The individual spectrum at each focus is retrieved and reconstructed from the superimposed spectra of the multifocal patterns. We demonstrate that the multifocal approach improves both the signal-to-noise ratio (SNR) and the imaging speed of Raman microscopy. Compared to the traditional point scan, at optimal detector conditions, the multifocal approach can be two-times faster for achieving the same image quality and SNR, or provides spectra with three-times higher SNR while applying the same energy dose at the focus. Such improvements of imaging speed and SNR increase up to one or two orders of magnitude under higher noise conditions, such as higher readout rate and higher detector temperatures. The multifocal approach presents advantages for certain imaging situations, such as when heating related damage limits the excitation energy dose that can be applied to the sample.

Motions inside living tissues are ubiquitous signatures of the active processes involved in the maintenance of cellular function and health. An innovative label-free non-invasive imaging technology called Biodynamic Imaging (BDI), based on low-coherence interferometry and digital holography, captures intracellular dynamics through ultra-low-frequency Doppler spectra encoding a broad range of cellular and subcellular motions. Clinical trials of BDI assessing patient response to therapy are underway in human ovarian, esophageal and breast cancers.

By virtue of its chemical specificity and label-free nature, Raman spectroscopy is a ubiquitous tool in analytical chemistry which has recently found increasing interest in biology and medicine, as technical improvements steadily enable applications. In particular, the low-frequency region of the Raman spectrum, which contains rich information about intermolecular interactions and higher-order structure, has promise for biological applications. However, detection of low-frequency modes remains a challenge with conventional techniques for Raman spectroscopy. Here, we propose and experimentally demonstrate Sagnac-enhanced impulsive stimulated Raman scattering (SE-ISRS), a method for time-domain Raman spectroscopy that provides highly sensitive low-frequency Raman spectra at all probe frequencies.

Altered biomechanics and phase transitions are implicated as key photogenic triggers in neurodegenerative diseases. However, standard methods to measure them require invasive contact with the sample or provide low spatial resolution. Here, we demonstrate Brillouin microscopy as a potential tool to investigate liquid-to-solid phase transitions in intracellular compartments in response to expression of ALS-linked proteins. In particular, we show how intracellular stress granules exhibit altered biomechanics in response to recruitment of RNA-binding proteins, such as ELAVL4 and FUS. Results pave the way to a better understanding of the dynamics that lead to formation of solid aggregates during neurodegeneration.

Autofluorescence spectroscopy can provide information on the metabolic status of cellular systems, but extensions of these techniques to turbid media such as tissues is complicated by the presence of multiple scattering, background fluorescence, and intrinsic absorption. Phasor analysis is a class of analytical approaches for the real-time assessment of emission signals that could be used to decipher cellular-level metabolic status of tissues. Spectral phasor analysis was originally developed for the rapid segmentation of hyperspectral images and has since been used for monitoring cellular NAD(P)H conformation from UV-excited cellular autofluorescence. Specifically, we showed previously that chemically induced autofluorescence responses in Saccharomyces cerevisiae (baker’s yeast) suspensions could not be accounted for using the two-component free vs. protein-bound model for conformation. Rather, by considering a series of physically similar and dissimilar chemicals acting on multiple metabolic pathways, we showed that responses affecting different pathways, e.g., involving cellular respiration versus oxidative stress, could be distinguished. Here, we seek to extend this pathway-level interpretation to the sensing of cellular metabolism in tissues by monitoring the cyanide-induced metabolic response of yeast cells embedded in media containing 9-cyanoanthracene or collagen as sources of background emission. Despite the similarity between autofluorescence and background spectra, we observe spectral behavior consistent with the discrimination of the metabolic response from the background emission. Performance over specifically selected noncontinuous spectral bands to reject chromophore absorption is also assessed.

Label-free high-resolution visualization of Alzheimer’s disease (AD) neuropathological hallmark, amyloid β (Aβ) plaques, is one of the prime goals of neuroscience. Till today, traditional histological procedures, which rely on fixation and tedious staining of tissues, can only provide definitive confirmation of AD. However, recent studies have shown that label-free third harmonic generation (THG) microscopy, a virtual transition based technology, can provide structural information of biological tissues with subcellular 3D resolution. In this study, using a 1263 nm Cr: Forsterite laser source, we performed THG studies on 3xTg AD mice brain tissues in vitro, with a focus on contrast origin evaluation for plaques. Our THG study can clearly differentiate, with very high resolution, neuropathological hallmark of AD: Aβ plaques. Moreover, THG can also distinguish white and gray matter along with axons, and soma of brain. The origin of THG contrasts for various structures of brain including AD pathological hallmarks were verified through standard immunohistochemical staining procedures. Our preliminary study has successfully demonstrated the capability of THG in revealing AD histopathological features with sub-femtoliter resolution without the need of any exogenous staining of the tissues.

Multispectral assessment of cell autofluorescence gives a direct window into the molecular processes occurring within those cells. This can be used to non-invasively characterise and classify various cellular properties without requiring fixation, dyes or transformation. Human mesenchymal stem/ stromal cells (MSCs) have great potential to contribute to regenerative medicine, especially with regards to autologous transplantation. However, this capacity is often limited by inherent properties of cell lines, which prevent their being sufficiently expanded after derivation for effective clinical application. The investigation of these properties requires numerous, time and labour-intensive assays. In this study we have used correlative microscopy based on multispectral images of cell autofluorescence then correlated to functional assays in order to construct multispectral signatures of numerous inherent cell characteristics. These included cell cycle status (indicating the proportion of cells undergoing cell division at a given time), cell ‘age’ (number of passages undergone, indicating capacity for further expansion), and β- galactosidase (a marker of senescence, indicating cells which can no longer divide). This study has established a single protocol, in place of multi-functional assays, to characterize the growth and differentiation capacity of hMSC lines using a non-invasive approach.

Tissue morpho-mechanics is gaining an increasing relevance in various fields, including biology, medicine, pathology, tissue engineering, and regenerative medicine, since it targets the relationship between morphological features and mechanical properties in biological tissues, which plays an important role in various biological processes including metastasis, wound healing and tissue regeneration. In particular, in every biological tissue, morphological, biochemical and mechanical properties are tightly connected and they influence each other in a correlative manner. For this reason, a correlative approach employing multiple techniques is ideal for targeting tissue morpho-mechanics with an optical approach. Here we report a correlative study performed by optical microscopies, disclosing the supramolecular collagen morphology correlated with its biomechanical and biochemical analyses. In particular, using human corneal tissue as a benchmark, we correlate Second-Harmonic Generation maps with mechanical and biochemical imaging obtained by Brillouin and Raman micro-spectroscopy, demonstrating that the peculiar mechanical functionality of so-called sutural lamellae originates from their distinctive supramolecular organization. A theoretical model based on the ultrastructural symmetry of corneal lamellar domains provides the interpretation of the experimental data at the molecular scale. The proposed methodology opens the way to the non-invasive assessment of tissue morpho-mechanics and holds the potential to be applicable to a broad range of biological and synthetic materials.

Automated, unbiased methods of non-invasive cell monitoring able to deal with complex biological heterogeneity are fundamentally important for biomedical research. Label-free imaging provides information about endogenous autofluorescent metabolites, enzymes and cofactors. Our multispectral fluorescence imaging technique allows precise quantification of native fluorophores in cells and tissues. This study uses label-free multispectral analysis to extract different fluorophores and redox ratio from single cells (oocytes, cultured cancer cells) as well as blastocyst embryos. Additionally, we characterise the molecular composition, structure and functional status of ex vivo healthy bovine and osteoarthritic human knee articular cartilage to assess 2 types of experimental treatments.

Emerging optical imaging techniques such as hyperspectral imaging (HSI) provide a promising non-invasive solution for intraoperative tissue characterisation with the potential to provide rich tissue-differentiation information over the entire surgical field. Neuro-oncology surgery would especially benefit from detailed real-time in vivo tissue characterisation, improving the accuracy with which boundaries of safe surgical resection are delineated and thereby improving patient outcomes. Current systems are limited by challenges with processing the HSI data because of incomplete characterisation of the optical properties of tissue across the complete visible and near-infrared wavelength spectrum. In this study, we characterised the optical properties of various freshly-excised brain tumours and normal cadaveric human brain tissue using a dual-beam integrating sphere spectrophotometer and the inverse adding-doubling technique. We adapted an integrating sphere to analyse 2 mm-thick tissue samples measuring 4 – 7 mm in diameter and validated the experimental setup with a tissue-mimicking optical phantom. We investigated the different spectral signatures of freshly-excised tumour tissues including pituitary adenoma, meningioma and vestibular schwannoma and compared these to normal grey and white matter, pons, pituitary, dura and cranial nerve tissues across the wavelength range of 400 – 1800 nm. It was found that brain and tumour tissues could be differentiated by their optical properties but the freezing process did alter the tissues’ relative absorption and reduced scattering coefficients. In this work, we have demonstrated a method to characterise the optical properties of small human brain and tumour specimens that may be used as a reference dataset for developing optical imaging techniques.

The pathological examination of regulatory relevant tissue samples is done manually and highly time consuming. Label free optical technologies like quantitative phase microscopy (QPM) and Optical Coherence Tomography (OCT) provide fast analysis of fixated cells and tissues by biophysical parameters. Combination of various wavelengths allows label free QPM tissue characterization equivalent to HE staining. Recent OCT studies showed 3D imaging of rat embryo details as bone structures, full vertebra, whole feet and parts of the skull. The multi-functional potential of QPM and OCT represents a promising new approach for label free phenotyping of regulatory relevant cell and tissue alterations.

Optical coherence tomography (OCT) suffers from speckle noise due to use of spatially coherent light, leading to significant reductions in image quality. Angular compounding techniques have been applied to remove speckle noise, but, existing image registration methods only guarantee pure angular compounding in the focal plane and produce spatial averaging in the defocused regions. This work develops an image registration model to correctly localize the real-space location of every pixel in OCT images captured at multiple incident angles, with digital focusing to further increase the lateral resolution and contrast. In the composite image, speckle suppression and resolution enhancement are realized.

We demonstrate wide-field intraoperative in vivo PS-OCT and OCT-A imaging of non-human primate (NHP) peripheral nerves using a portable OCT prototype. Imaging was performed on healthy and surgically repaired (12 months after graft surgery) radial nerves in upper extremity of rhesus macaques (n=17) over surgically exposed nerve segments of up to 8 cm in length. We describe the capabilities and challenges of PS-OCT and OCT-A in assessing NHP peripheral nerves. Further, guided by imaging of NHP peripheral nerves, we discuss the required advancements in core polarimetry techniques to enhance the prospects of clinical utility of PS-OCT in peripheral nerve imaging.

In this work we develop a theoretical framework for analyzing generic speckle reduction algorithms including all possible light modulations: angular compounding, spatial light modulators and others. We show that speckle reduction using Laguerre Gauss modes - LG Speckle Reduction achieves the theoretical boundary of speckle reduction using minimal number of samples and therefore is optimal. We demonstrate LG Speckle reduction in in-vitro sample and show that speckle reduction is in line with our theoretical model. We further develop LG Speckle Reduction to show that it can outperform speckle reduction outside of the focal plane by most speckle reduction methods and generates fast and high-quality images.

Direct visualization and characterization of nanosized biological particles such as viruses, vesicles and protein aggregates are important for various applications in medicine. Specifically, exosomes (30-150 nm in diameter) gained huge interest due to their potential role as a biomarker in cancer diagnosis and prognosis, however direct detection of these particles is challenging due to their small size. Interferometric microscopy allows detection of these particles without using any labels. We show that visibility of nanoparticles can be enhanced in interferometric microscopy by utilizing defocused images. In this paper, with the proposed method Depth Scanning Correlation, we demonstrate label-free detection of individual exosomes isolated from breast cancer cell culture isolated by using Exosome Total Isolation Chip (ExoTIC). Proposed imaging system combined with an isolation tool, can be used in a wide range of applications, where label-free detection of single biological nanoparticles is needed.

An optical biosensor is a compact analytical device formed by a bio-recognition sensing element integrated to an optical transducer system which translates a signal into a readable outcome that is measured by the detector. The target analyte interacts with an immobile bio-recognition element giving rise to a signal proportionate to the concentration of a measured analyte. Optical biosensors offer great advantages over conventional analytical techniques. Specifically, they can provide multiple capabilities such as user-friendly operation, real-time analysis, rapid response, high sensitivity and specificity, portability, label-free detection and cost-effectiveness. As a result, they possess suitable features critical for point-of-care diagnostics. In this study, a home-build surface plasmon resonance (SPR) optical biosensor device was used to analyse interactions between the bio-recognition sensing element and an analyte on the biosensing layer. The transducer consisted of silica dioxide (SiO₂) substrate layer where a thin layer of gold was deposited. Mycolic acid antigens from mycobacterium tuberculosis (bovine strain) were immobilised on the biosensing layer and used as biorecognition sensing elements to capture tuberculosis (TB) antibodies (analyte). From our findings, it was realised that the mycolic acid successfully captured TB antibodies resulting in a detectable signal which paves a way for the development of the point-of-care device.

Organophosphates (OPs) are a class of pesticides and chemical warfare, several of which are highly toxic. Based on the newest report, there are nearly three million poisonings per year resulting in two hundred thousand deaths around the world. Inspired by the Liquid Crystals (LCs)’ extraordinary properties, we employed a customized LCs sensor to detect OPs with high sensitivity & selectivity, fast response time, and reusability. As proof of our research, four different measurements were operated in the experiments: sensitivity, response time, selectivity, and reusability. The results showed that the limit of detection (LOD) of our proposed sensing method can selectively detect OPs lower to the 10ppb concentration in 25 seconds in four different vapors (OPs, Steam, Ethyl alcohol, and Methylbenzene), which was far lower than the US EPA safety levels for OPs and also faster than the other competitors. Further, the sensor can be recovered by the nitrogen gas treatment in 30 seconds.

In the context of biosensing and plasmonic imaging, we will report on actual characterization of structured biological objects such as layers of cell films. To achieve this, compared to conventional plasmonic systems, one can optimize not only the optical system but also the substrates that support the plasmonic modes and diminish the propagation length associated to this non-local imaging modality. We will focus on this latter point and present both simulation and experimental results, making use of flat and nanostructured geometries that allow achieving better imaging resolution yet keeping good contrast.

Collagen fiber architecture plays an important role in the mechanical properties of soft tissues. Conventional polarized light microscopy done using linear polarizers and, sometimes, quarter-wave plates is a label-free imaging technique for quantifying collagen fiber architecture, specifically distribution and orientation. However, this technique has several limitations. First, it requires acquiring multiple images with different polarization states, which precludes many timesensitive applications. Second, post-processing, especially image registration, reduces the level of detail discernible. Third, the added optical elements may cause glare under coaxial illumination, thus complicating the use of reflected light microscopy. We have recently demonstrated instant polarized light microscopy (IPOL), that requires only one image and therefore no registration. IPOL utilizes wavelength-dependent polarization to modify the spectrum of the illumination, generating visible colors that depend on fiber orientation and density. Herein we present two further advances on IPOL: we extend it to work with coaxial illumination allowing transmitted and reflected light microscopy, and we integrate it in a dissecting microscope. This permits real-time imaging, limited only by the camera frame rate, making it possible to track dynamic events, such as fast-acting responses to external forces or moving objects. We demonstrate IPOL with a field of view of 11 mm and a long working distance of 65 mm, which simplifies testing of large samples. IPOL provides both fiber distribution and orientation information in a single true-color snapshot, and therefore, it is suitable for time-sensitive applications.

Graded index (GRIN) lenses focus light through a radially symmetric refractive index profile. It is not widely appreciated that the ion-exchange process that creates the index profile also causes a radially symmetric birefringence variation. This property is usually considered a nuisance, such that manufacturing processes are optimized to keep it to a minimum. Here, a new Mueller matrix (MM) polarimeter based on a spatially engineered polarization state generating array and GRIN lens cascade for measuring the MM of a region of a sample in a single-shot is presented. We explore using the GRIN lens cascade for a functional analyzer to calculate multiple Stokes vectors and the MM of the target in a snapshot. A designed validation sample is used to test the reliability of this polarimeter. To understand more potential biomedical applications, human breast ductal carcinoma slides at two pathological progression stages are detected by this polarimeter. The MM polar decomposition parameters then can be calculated from the measured MMs, and quantitatively compared with the equivalent data sampled by a MM microscope. The results indicate that the polarimeter and the measured polarization parameters are capable of differentiating the healthy and carcinoma status of human breast tissue efficiently. It has potential to act as a polarization detected fiber-based probe to assist further minimally invasive clinical diagnosis.

Mueller polarimetry is a powerful characterization technique for a variety of samples and a promising optical-biopsy tool for early detection of cancer. Recent advances in Mueller imaging devices allow the collection of large ex-vivo and invivo image databases. Although the technique is sensitive to subtle changes in the micro-organization of tissue, the Mueller matrices of such complex media contain intertwined polarimetric effects and are difficult to interpret. To identify the polarimetric signature of a given tissue modification (cancerous or not), machine learning tools are particularly well suited. However, a statistically sound approach is needed to make the most out of these tools and avoid common pitfalls. We present a global statistical framework based on decision theory. It consists of a complete preprocessing and analysis pipeline for polarimetric bioimages. In the analysis stage, we use a loss-risk-based approach to automatically select the optimal classifier among a library of classifiers. The approach allows to determine the subset of polarimetric parameters of interest, to determine the parameters of the classifiers and to assess classifier performance using cross-validation. The proposed framework is illustrated with precancer detection on human ex-vivo cervical samples.

Tissue-based diagnostic relies on histological analysis of tissue cuts by a pathologist who links data derived from microscopy image content to a specific disease. We explored the potential of Mueller microscopy combined with density-based spatial clustering algorithm for applications with noise for the automated analysis of tissue histological cuts. Mueller microscopic images of unstained artificial human skin histological cuts were analyzed by applying differential Mueller matrix decomposition. Scalar retardance, depolarization and total transmitted intensity images were used for the automated segmentation of microscopic images. An algorithm of polarimetric data post-processing which mitigates the impact of tissue thickness fluctuations was suggested.

Most of the recent super-resolution imaging techniques developed for biology are based on fluorescence properties which generally require toxic dyes. We show how super-resolution can be obtained without dyes by simply adding 20-micrometer-diameter dielectric spheres on the sample under a microscope objective. The microsphere behaves as a non-classical lens, collecting evanescent waves in a full-field imaging process. Resolutions of up to between /5 and /7 can be reached in air and in immersion, respectively. For translucent biological samples, a dark-field optical setup is proposed. Performance of the label-free super-resolution technique is demonstrated through the imaging of fixed nerve sections from a mouse embryo brain.

Facile analysis of the architecture of the mammalian brain is key to understanding how brain function emerges during development and dysregulated in disorders including neurodegeneration. Immunolabeling of mammalian brain tissue, especially scarce human brain tissue, is time-consuming, can introduce sample-to-sample variation, and is not compatible with live imaging. We report joint optimization of polarization-resolved label-free imaging and deep learning to map brain architecture. We visualize diverse structures in human brain tissue by mapping optical properties of density, birefringence, orientation, and scattering. We design computationally efficient variants of U-Nets to predict tract distribution and cell types from intrinsic optical properties of the tissue.

The recent development and integration of multiple non-linear microscopy techniques in a single instrument has provided new opportunities for integrating morphological and functional information and for correlating the observed molecular and cellular changes with disease behaviour. In particular, multimodal non-linear/linear imaging is able to perform a morpho-chemical quantitative analysis in tumour cells and tissue specimens, providing a high-resolution label-free alternative to both histological and immune-histochemical examination of tissues. In this talk, a brief overview on the non linear and linear laser brain imaging techniques will be displayed for label free detection. Morpho-chemical characterization of tissue will be displayed as an interesting tool for early diagnosis of pathologies: different kind of approaches will be shown for in vivo imaging assisted surgery operation or as tools to support anatomo-pathologists decision.

Heart failure (HF) has a large impact on patient outcomes and health care costs. Objective monitoring of the pitting edema level of a HF patient may help clinicians reduce the amount of readmissions. ChemImage is developing a Molecular Chemical Imaging (MCI) device for monitoring HF patients that will non-invasively quantify peripheral edema. Results from a completed in-human clinical trial will be presented demonstrating ability to discriminate between healthy volunteers and HF patients with all levels of pitting edema and correct prediction of peripheral edema grade across the patient population. Follow-on clinical trials will address monitoring patients during treatment.

Chemical microscopy utilizing fingerprint spectroscopic signals is able to map the chemical contents temporally and spatially. Such capacity opens a new window to visualize the orchestra of molecules and/or biological structures inside living systems. Cheng and his research team have been dedicated to pushing the boundary of chemical microscopy in the entire spectrum of molecular spectroscopy, discovering molecular signatures in diseases, and translating label-free techniques to clinic for molecule-based precision diagnosis or treatment.

We present coherent anti-Stokes Raman (CARS) imaging with our robust all-fiber optical parametric oscillator (FOPO) that is tunable across the whole bio-relevant spectral region (700 to 3530 wavenumbers) in less than 5 ms. The all-fiber operation allows to operate the FOPO on a regular laboratory cart, maximize temporal pulse overlap by dispersion management and inherently minimize beam-pointing. We present first CARS images of DMSO-d6 infiltration into mouse skin imaged in the silent as well as CH-stretch region without any alignment other than electronically tuning the wavelength. The FOPO is ideally suited for label-free, hyperspectral imaging, e.g., in translational research.

Coherent Raman microspectroscopy utilizing either stimulated Raman scattering or coherent Raman scattering is gaining popularity in the field of biological imaging. It is generally assumed that coherent Raman spectroscopy provides a stronger signal than conventional spectroscopy based on spontaneous scattering, allowing much faster imaging. In this report we examine non-trivial advantages of coherent Raman microspectroscopy, which haven’t been given significant attention lately. In particular, we demonstrate the potential of improving sensitivity of Raman spectroscopy and ability to sense the local nanoenvironment.

We present a novel design for a Stimulated Raman Scattering (SRS) microscope based on a dual beam femtosecond laser in combination with spectral shaping through a fast and narrowband Acousto Optical Tunable Filter. This configuration allows the measurement of broad SRS spectra, all the way from fingerprint region to CH stretch region without any modification of the optical setup. High spectral resolution over a broad spectral region allows label-free quantitative imaging of biological samples. We will show the application of our SRS system to a quantitative study of lipid droplets in colon Cancer Stem Cells.

We present a flexible, compact, and multimodal nonlinear endoscope (2.2 mm outer diameter) based on a resonantly scanned negative curvature hollow-core double-clad fiber. The fiber design allows distortion-less, background-free delivery of femtosecond and picosecond excitation pulses and the back-collection of nonlinear signals through the same fiber. Sub-micron spatial resolution together with large field of view is made possible by the combination of a miniature objective lens together with a silica microsphere lens inserted into the fiber core. We demonstrate coherent anti-Stokes Raman scattering, 2-photon fluorescence and second harmonic generation imaging at a rate of 10 frames/s.

Iron is an essential element required for human life, and is highly regulated in the body. Iron deficiency leads to many adverse health effects, such as anemias. The exact mechanisms of iron release in cells are not well known. We developed a Raman micro-spectroscopy technique that allows detection of transferrin (Tf) bound iron inside intact human cells. Ferric iron (Fe³⁺) bound to serum Tf is internalized into cells via the transferrin receptor (TfR). Methods that allow determining when and where Tf releases iron inside a cell lead to a better understanding of disease progression, including cancer. We have previously shown that Raman micro-spectroscopy is able to detect and quantify the Tf-bound iron in human breast cancer T47D cells. In this work, we applied hyperspectral Raman imaging to visualize the spatial distribution of Tf-bound iron in human breast cancer T47D and MDAMB231 cells internalized with iron-loaded Tf. We have also shown that Raman imaging can quantify the amount of iron under different times of Tf internalization prior to fixation. Raman microspectroscopy provides a unique way to determine the amount of iron under different Tf internalization times by employing the Raman metric, which was used to quantify iron content in iron bound Tf (holo-Tf) samples. Importantly, Raman microspectroscopy can be used to follow iron release from Tf in breast cancer cells. Determining the kinetics and location of iron release in cancer cells is key to further our understanding of iron metabolism during cancer progression.

Hirschsprung disease (HD) is a congenital disorder in the distal intestine and is characterized by the absence of nerve ganglion cells (aganglionosis). HD affects newborns by causing severe constipation. Surgical management is needed and consists of the accurate identification and removal of the aganglionic segment and the reconstruction of the intestinal tract. The gold standard for the definitive diagnosis of the aganglionic segment is the histologic evaluation of rectal biopsies through stained sections. However, it is a time-consuming procedure, and recognized factors for inaccurate diagnoses have been commonly reported. In recent years there has been much interest in the use of optical techniques to improve diagnostics in health care. Here, for improving the identification of ganglion cells, we propose an ex-vivo study to evaluate a combination of label-free optical modalities: second harmonic generation (SHG); two-photon autofluorescence; and Raman spectroscopy. SHG and autofluorescence images have been used to locate regions of interest in the tissue for Raman analysis, which acquires a molecular fingerprint of the ganglion cells, without needing any stains or labels. Multivariate statistical analyses of the Raman spectral data have been used for objective identification of the ganglion cells in the tissue samples with high accuracy.

Extracellular vesicles (EVs) are biologically derived nanovectors important for intercellular communication and trafficking, yet understanding of their underlying biological mechanisms remains poor. Advances have been hampered by both the complex biological origins of EVs and a lack of suitable imaging techniques. Here, we present a strategy for simultaneous in vitro imaging and molecular characterisation of EVs in 2D and 3D based on Raman spectroscopy and minimally-obstructive metabolic deuterium labelling. Metabolically-incorporated deuterium acts as a bio-orthogonal Raman-active tag for direct Raman identification of EVs and provides insights into their biocomposition and trafficking, with implications for their development as therapeutic delivery vectors.

Fourier Ptychography is a computational imaging technique able to decouple high resolution from wide field of view, bypassing the diffraction limit of the microscope. Since it does not rely on high precision mechanics or fluorescent imaging, it is of practical interest for implementation in low scale devices. Despite its gains, realizing a functional low-cost setup working at the theoretical limits is challenging due to many factors causing discrepancies between theory and practice. Misalignment of the light emitting diode array (LED-array), optical system aberrations and use of partial coherent sources are common issues which have been addressed with calibration algorithms. However, physical interpretation of how these factors influence the algorithm and cause mismatches between theory and practice has had little attention so far. This work provides a discussion based on simulation results on the effect of the partial coherence of the source. From obtained results, an optimal set of LEDs for data acquisition is described which avoids degeneracy caused by partial coherence and is based on the numerical aperture (NA) of the objective and source parameters such as bandwidth and size.

Six-pack holography (6PH) is the first holographic imaging modality that is more spatial bandwidth efficient than on-axis holography. 6PH utilizes a low-coherence light source and a phase delay plate to enable the spatial multiplexing of six off-axis holograms without cross-talk between the six sample and six reference beams, and is thus capable of acquiring six off-axis holograms in a single camera exposure. We applied 6PH to synthetic aperture (SA) superresolution, and produced an SA capable of dynamically increasing the resolution by a factor of 1.62. 6PH presents a valuable general purpose technique for all holography applications and label-free imaging.

Deep learning is a class of machine learning techniques that uses multi-layered artificial neural networks for automated analysis of signals or data. The name comes from the general structure of deep neural networks, which consist of several layers of artificial neurons, each performing a nonlinear operation, stacked over each other. Beyond its main stream applications such as the recognition and labeling of specific features in images, deep learning holds numerous opportunities for revolutionizing image formation, reconstruction and sensing fields. In this presentation, I will provide an overview of some of our recent work on the use of deep neural networks in advancing holographic microscopy and sensing systems, also covering their biomedical applications.

Digital magnitude image rendering in Gabor holography can be performed by a convolutional neural network trained with a fully synthetic database formed by image pairs generated randomly. These pairs are linked by a numerical model propagation of a scalar wave field from the object to the sensor array. The synthetic database is formed by generating images made from source points at random locations with random brightness on a black background. Successful prediction of experimental Gabor holograms of microscopic worms by a UNet trained with 50,000 random image pairs is achieved, and a classifier-based regularization for twin-image removal is investigated.

We report the full-field imaging of the mechanical deformations accompanying the action potential in primary cortical neurons using ultrafast quantitative phase imaging (QPI) with a temporal resolution of 0.1 ms and a membrane displacement sensitivity of <0.2 nm per pixel. The average displacements were ~0.7 nm on cell somas and ~0.5 nm on neurites. Finite element modeling based on the 3D shape extracted from confocal imaging and on scaling of the surface tension with trans-membrane voltage yielded the deformation map during action potential, which matched the features of the experimental results, including the displacement amplitude, time course, and spatial distribution.

We apply label-free imaging using digital holographic microscopy to analyze different cancer cell lines. Separation of cell lines based on extraction of amplitude and phase map variations along with post-processed, population specific parameters, was accomplished using machine learning. These data are used to train a neural network algorithm that attains accurate discrimination of non-adherent cancer cells.

Progression of acute diseases such as cancer often depend on a small number of specialized cells. Identification and isolation of the cells intact is of utmost importance. Label-based approaches are incompatible with noninvasive cell processing. We propose a a new label-free technique for noninvasive and automated cell processing and discrimination of individual cells. By acquiring holograms of each cell and achieving its optical path delay (OPD) profile, we extract features that highly differentiate cancerous cells from heterogeneous blood sample. Using a dielectrophoresis flow chamber to control each cell, we can than collect the cancerous cells for further analysis and classification.

Quantitative phase microscopy (QPM) is a label-free imaging technique to quantify various biophysical parameters, such as refractive index, optical thickness, cell dry mass, and dynamic membrane fluctuations. Accurate determination of these parameters requires the use of a QPM system with high temporal phase stability and high spatial phase sensitivity. We report a QPM system based on a common-path interferometer with high temporal phase stability and high spatial phase sensitivity. The proposed QPM system is highly temporally stable, compact and easy to align and implement. The interference pattern can be obtained quickly even with a low coherent light source. In order to realize high spatial phase sensitivity, we used partially spatially coherent (pseudo-thermal) light source for illumination. Due to the partial spatial coherent nature of the light source, a speckle-free interferogram/hologram is recorded over the entire field-of-view. Two types of speckle free QPM systems are implemented using common path Fresnel biprism as well as lateral shearing interferometers. A Fresnel biprism is used in the self-referencing mode, thus offering the advantage of no optical power loss in addition to high temporal stability and the least speckle artifacts. Furthermore, it is very easy to implement, as the system completely replaces the need for spatial filtering at the source end as well as for the reference beam generation. In another configuration, we used a lateral shearing interferometer. The scattered light from the object is collected by the microscope objective lens and passes through a 4mm thick optically flat parallel plate to generate the interference pattern. Phase maps of human RBCs are reconstructed and the results are compared for fully and partially coherent light illumination.

We demonstrate rapid, sensitive and label-free detection of whole-cell E. coli utilizing interferometric reflectance imaging enhancement technique, with a limit of detection of 2.2 CFU/ml from a buffer solution with no sample preparation. Our sensor platform provides visualizing and counting individual pathogens captured on the large sensor surface, as well as morphological characterization of the captured particles by high optical magnification imaging modality for validating the recorded detection events as the target bacteria. In addition, we show that our biosensor's detection capability is unaffected by the sample complexity by testing its performance in tap water. Also, the specificity of our biosensor is validated by comparing its response to target bacteria E. coli and non-target bacteria S. aureus, K. pneumonia and P. aeruginosa. Therefore, our sensitive and label-free detection method offers new perspectives for direct bacterial detection in real matrices and clinical sample.

Quantitative phase imaging (QPI) is an important tool in biomedicine that yields label-free access to cellular and subcellular structures with nanometer-scale sensitivity. However, implementation of QPI involves a transmissionbased geometry and requires thin samples, severely hindering its overall utility in biomedicine. Here we describe our recently developed method, quantitative oblique back illumination microscopy (qOBM), which overcomes this significant limitation and achieves epi-mode, tomographic, quantitative phase imaging in thick samples, including intact thick tissues. Here we describe the method in detail, show validation experiments and results from thick scattering samples.

Avoiding adverse effects of staining reagents on cellular viability and cell signaling, label-free cell imaging and analysis is essential to personalized genomics, drug development, and cancer diagnostics. By analyzing the images of cells, imagebased cell analytic methodologies offer a relatively simple and economical way to understand the cell heterogeneities and developments. Owing to the developments in high-resolution image sensors and high-performance computation processors, the emerging lens-less digital holography techniques enable a simple and cost-effective approach to obtain label-free cell images with large field of view and microscopic spatial resolution. In this work, the lens-less digital holography technique is adopted for image-based cell analysis. The holograms of three kinds of cells which are MDA-MB231, EC-109 and MCF-10A respectively were recorded by a lens-less digital holography system composed of a laser diode, a sample holder, a sensor and a laptop computer. The acquired holograms are first high-pass filtered. Then the amplitude images were reconstructed using the angular spectrum method and the sample to sensor distance was determined using the autofocusing criteria based on the sparsity of image edges and corner points. The convolutional neural network (CNN) was used to classify the cells. The experiments show that an accuracy of 97.2% can be achieve for two type cell classification and 91.2% for three type cell classification. It is believed that the lens-less holography combining with machine learning holds great promise in the application of stainless cell imaging and classification.

In order to advance quantitative phase microscopy as a significant clinical tool, we have implemented a new modality, holographic cytometry, which provides high resolution phase maps at high frame rates using several novel advances such as high speed line scan cameras integrated with microfluidic and illuminated with pulsed light source. The system is used to examine carcinogenic changes in epithelial cells which have been exposed to heavy metals in population sizes that are diagnostically relevant.

Quantitative phase imaging (QPI) provides a label free method for imaging live cells and allows quantitative estimates of cell volume. Because the phase of light is not directly measurable at an imaging sensor, QPI techniques involve both hardware and software steps to reconstruct the phase. Digital holographic microscopy (DHM) is a QPI technique that utilizes an interferometer to combine a reference beam with a beam that passes through a specimen. This produces an interference pattern on the image sensor, and the specimen’s phase can be reconstructed using diffraction algorithms. One limitation of DHM is that the images are subject to coherent diffraction artifacts. Transport of intensity (TIE) method, on the other hand, uses the fact that defocused images of a specimen depend on the specimen’s phase to determine the phase from two or more defocused images. Its benefit over DHM is that it is compatible with conventional bright field imaging using sources of relatively low coherence. Although QPI methods can be compared on a variety of static phase targets, these largely consist of phase steps rather than the phase gradients present across cells. In order to compare the QPI methods described above on live cells, rapid switching between QPI modalities is required. We present results comparing DHM and TIE on a custom-built microscope system that allows both techniques to be used on the same cells in rapid succession, which allows the comparison of the accuracy of both measurements.

We examine multiphoton-produced optical signals waveguided through single ZnO nanorod (NR) using a newly developed, scanning offset-emission hyperspectral microscopy (SOHM) technique. SOHM acquires spectrally indexed and spatially resolved intensity maps/spectra of waveguided light intensity, while excitation/emission collection positions and light polarization are scanned. Hence, the powerful measurement capabilities of SOHM enable quantitative analyses of the different ZnO NR waveguiding behaviors specific to the multiphoton-generated emissions as a function of measurement position, and the optical origin of the guided signal. We subsequently reveal the distinct waveguiding behaviors of single ZnO NRs pertaining to the variously originated signals and discuss particularly attractive ZnO NR properties in CARS waveguiding. In this talk, I will present the distinctive CARS waveguiding nature through single ZnO NR, exhibiting high position and polarization-dependence.

In this work, we studied the nature of the molecular bonds involved in the red blood cell aggregation process by the coherent anti-Stokes Raman spectroscopy technique. Images were acquired with a commercial Leica TCS SP8 CARS confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany) temporally and spatially overlapping the pulses of two sources in the focal plane of the microscope. A pump wavelength of 810 nm to 817 nm was used for the CARS mode simultaneously with the Stokes beam at 1064 nm to excite the vibratory resonance of the symmetric hydrocarbon bonds in the lipids and that of the bonds in amino acids of the proteins. The Raman shift was also observed at the 1200 cm⁻¹ range to study possible variations in the sialic acid on the cell membrane produced by concentrations of dextran 500 in the suspension medium. Curves of lifetime emission distribution were obtained for untreated erythrocytes and treated erythrocytes with dextran 500, particularly at a pump wavelength of 904 nm.

During prostate cancer progression, cancerous epithelial cells can undergo epithelial-mesenchymal transition (EMT). EMT is a crucial mechanism for the invasion and metastasis of epithelial tumors characterized by the loss of cell-cell adhesion and increased cell mobility. It is associated with biochemical changes such as epithelial cell markers Ecadherin and occludins being down-regulated, and mesenchymal markers vimentin and N-cadherin being upregulated. These changes in protein expression, specifically in the cell membrane, may be monitored via biophysical principles, such as changes in the refractive index (RI) of the cell membrane. In our previous research, we demonstrated the feasibility of using cellular RI as a unique contrast parameter to accomplish label-free detection of prostate cancer cells. In this paper, we report the use of our Photonic-Crystal biosensor in a Total-Internal-Reflection (PC-TIR) configuration to construct a label-free biosensing system, which allows for ultra-sensitive quantification of the changes in cellular RI due to EMT. We induced prostate cancer cells to undergo EMT by exposing these cells to soluble Transforming Growth Factor Beta 1 (TGF-β1). The biophysical characteristics of the cellular RI were quantified extensively in comparison to non-induced cancer cells. Our study shows promising clinical potential in utilizing the PC-TIR biosensing system not only to detect prostate cancer cells, but also to evaluate changes in prostate cancer cells due to EMT.

The detection of cardiac troponin I (cTnI) is clinically used to monitor myocardial infarctions (MI) and other heart diseases. The development of highly sensitive detection assays for cTnI is needed for the efficient diagnosis and monitoring of cTnI levels. Traditionally, enzyme-based immunoassays have been used for the detection of cTnI. However, the use of labelfree sensing techniques have the advantage of potentially higher speed and lower cost for the assays. We previously reported a Photonic Crystal-Total Internal Reflection (PC-TIR) biosensor for label-free quantification of cTnI. To further improve on this, we present a comparative study between an antibody based PC-TIR sensor that relies on recombinant protein G (RPG) for the proper orientation of anti-cTnI antibodies, and an aptamer-based PC-TIR sensor for improved sensitivity and performance. Both assays relied on the use of polyethylene glycol (PEG) linkers to facilitate the modification of the sensor surfaces with biorecognition elements and to provide fluidity of the sensing surface. The aptamer-based PC-TIR sensor was successfully able to detect 0.1 ng/mL of cTnI. For the antibody-based PC-TIR sensor, the combination of the fluidity of the PEG and the increased number of active antibodies allowed for an improvement in assay sensitivity with a low detection limit of 0.01 ng/mL. The developed assays showed good performance and potential to be applied for the detection of cTnI levels in clinical samples upon further development.

In the recent studies of cartilage imaging with nonlinear optical microscopy, we discovered that autofluorescence of chondrocytes provided useful information for the viability assessment of articular cartilage. However, one of the hurdles to apply this technology in research or clinical applications is the lack of image processing tools that can perform automated and cell-based analysis. In this report, we present our recent effort in the cell segmentation using deep learning algorithms with the second harmonic generation images. Two traditional segmentation methods, adaptive threshold, and watershed, were used to compare the outcomes of different methods. We found that deep learning algorithms did not show a significant advantage over the traditional methods. Once the cellular area is determined, the viability index is calculated as the intensity ratio between two autofluorescence channels in the cellular area. We found the viability index correlated well with the chondrocyte viability. Again, deep learning segmentation did not show a significant difference from the traditional segmentation methods in terms of the correlation.

There is a growing focus to adapt Polymerase Chain Reaction (PCR) to point-of-care (POC) testing to provide for a low-cost, rapid and reliable diagnostic instrument. Many studies proposed the integration of microfluidics with fluorophore-assisted or electrochemical amplicon detection methods to introduce a real-time miniature device for POC applications. However, their practicality in POC testing is limited due to their complex microfabrication, high cost, and intrinsic challenges due to their intercalation and hybridization-based detection. In this paper, we present a purely optical methodology without the addition of non-PCR reagents (electroactive or fluorogenic DNA intercalators) to enhance the reliability in quantitative PCR measurement of DNA yield. The determination of PCR results and DNA amplicon quantification are realized by monitoring transmitted power of a 260nm LED in PCR reaction at every thermal cycle. The least-square fits to transmission data demonstrate distinctive features to classify positive vs. negative PCRs and to quantify amplified products. This real-time UV monitoring system was combined with a VCSEL-based plasmonic thermocycler to accomplish fast amplification and detection in a simple and small-scaled footprint applicable for POC diagnostics.

Raman spectroscopy is commonly used for sample characterization in biology because vibrational information is very specific to the chemical bonds in molecules. This makes it an attractive approach for identification of biological materials such as toxins, viruses or even intact bacterial cells. In addition, Raman spectroscopy has a unique capability of providing label-free intrinsic chemical information, such as molecular bonds in living biological samples at tissue, cellular or subcellular resolution. However, Raman signals are weak and acquiring a spectrum with good signal to noise ratio requires long acquisition time. To overcome this disadvantage of low signal intensities from most biomolecules, enhancement effects are utilized. In this study, a home built Raman spectroscopy optical system combined with a gold thin film deposition was used to detect the HIV gp41 antibody. The Raman system makes use of 785 nm diode laser as excitation source and an Andor CCD camera as detection system. In addition, we report on Raman results obtained with HIV gp41 antiboby using a gold thin film deposition substrate. We could observe significant enhancement of Raman signal from the gold thin film layer deposition. These findings indicate the potential application of Raman spectroscopy in rapid biosensing detection.