This PDF file contains the front matter associated with SPIE Proceedings Volume 11966, including the Title Page, Copyright information, and Table of Contents.

Multidimensional fluorescence microscopy techniques produce dataset rich of information (space, emission spectrum and lifetime) to investigate photophysical processes in biological samples. To acquire a 4D dataset, one promising microscope design is based on the single-pixel camera scheme and on compressive sensing acquisitions, thanks to which the measurement time can be reduced. Within this framework, a computational step is required to move from the acquisition space to the pixel space and, subsequently, the analysis can be carried out exploiting the high dimensionality. In this work we present an experimental system and a fast-fit method that can produce a map of fluorophore concentrations in parallel to the measurement routine.

Fluorescence microscopy has been a significant tool to observe long-term imaging of embryos (in vivo) growth over time. However, cumulative exposure is phototoxic to such sensitive live samples. While techniques like light-sheet fluorescence microscopy (LSFM) allows for reduced exposure, it is not well suited for deep imaging models. Other computational techniques are computationally expensive and often lack restoration quality. To address this challenge, one can use various low-dosage imaging techniques that are developed to achieve the 3D volume reconstruction using a few slices in the axial direction (z-axis); however, they often lack restoration quality. Also, acquiring dense images (with small steps) in the axial direction is computationally expensive. To address this challenge, we present a compressive sensing (CS) based approach to fully reconstruct 3D volumes with the same signal-to-noise ratio (SNR) with less than half of the excitation dosage. We present the theory and experimentally validate the approach. To demonstrate our technique, we capture a 3D volume of the RFP labeled neurons in the zebrafish embryo spinal cord (30 μm thickness) with the axial sampling of 0.1 μm using a confocal microscope. From the results, we observe the CS-based approach achieves accurate 3D volume reconstruction from less than 20% of the entire stack optical sections. The developed CS-based methodology in this work can be easily applied to other deep imaging modalities such as two-photon and light-sheet microscopy, where reducing sample photo-toxicity is a critical challenge.

Cyclic AMP (cAMP) is a second messenger that regulates a wide variety of cellular functions. There is increasing evidence suggesting that signaling specificity is due in part to cAMP compartmentalization. In the last 15 years, development of cAMP-specific Förster resonance energy transfer (FRET) probes have allowed us to visualize spatial distributions of intracellular cAMP signals. The use of FRET-based sensors is not without its limitations, as FRET probes display low signal to noise ratio (SNR). Hyperspectral imaging and analysis approaches have, in part, allowed us to overcome these limitations by improving the SNR of FRET measurements. Here we demonstrate that the combination of hyperspectral imaging approaches, linear unmixing, and adaptive thresholding allow us to visualize regions of elevated cAMP (regions of interest – ROIs) in an unbiased manner. We transfected cDNA encoding the H188 FRET-based cAMP probe into pulmonary microvascular endothelial cells. Application of isoproterenol and prostaglandin E1 (PGE1) triggered complex cAMP responses. Spatial and temporal aspects of cAMP responses were quantified using an adaptive thresholding approach and compared between agonist treatment groups. Our data indicate that both the origination sites and spatial/temporal distributions of cAMP signals are agonist dependent in PMVECs. We are currently analyzing the data in order to better quantify the distribution of cAMP signals triggered by different agonists.

Wavefront-sensorless adaptive optics methods are often used to correct phase aberrations in optical systems and thus to improve imaging quality. However, sensorless methods have an intrinsic disadvantage of requiring multiple images that can lead to non-desirable photo-bleaching. We have proposed a machine learning assisted aberration correction method which could correct aberrations consisting of not fewer than five Zernike modes with as few as two images. We showed that our method could be used in microscopes to provide instant aberration predictions when imaging biological samples of non-specific structures. We showed that compared to conventional function fitting sensorless adaptive optics methods, the new method corrected much faster with observable advantages. This novel method has a great potential to be used in any adaptive optics equipped microscopy for efficient sensorless aberration correction for biomedical microscopy.

Hyperspectral imaging technologies (HSI) have undergone rapid development since their beginning stages. While original applications were in remote sensing, other uses include agriculture, food safety and medicine. HSI has shown great utility in fluorescence microscopy for detecting signatures from many fluorescent molecules; however, acquisitions speeds have been slow due to light losses associated with spectral filtering. Therefore, we designed a novel light emitting diode (LED)-based rapid excitation scanning hyperspectral imaging platform allowing users to obtain simultaneous measurements of fluorescent labels without compromising acquisition speeds. Previously, we reported our results of the optical ray trace simulations and the geometrical capability of designing a multifaceted mirror imaging system as an initial approach to combine light at many wavelengths. The design utilized LEDs and a multifaceted mirror array to combine light sources into a liquid light guide. The computational model was constructed using Monte Carlo optical ray software (TracePro, Lambda Research Corp.). Recent prototype validation results show that when compared to a commercial emission scanning spectral confocal microscope (Zeiss-LSM-980), the novel LED-based excitation scanning HSI prototype successfully detected and separated six fluorescent labels from a custom 6-label African green monkey kidney epithelial cells. We report on the prototype’s ability to overcome limitations of acquisition speeds, sensitivity, and specificity present in conventional systems. Future work will evaluate prototype’s light losses to determine latent design modifications needed to demonstrate the system’s feasibility as a promising solution for overcoming HSI acquisition speeds. This work was supported by NSF award MRI1725937.

Diffraction-limited imaging in microscopy is only possible if different layers within the objective’s working distance all have an uniform refractive index. However, in many practical imaging problems the samples are inhomogeneous in optical density, and refraction through them introduces field-dependent wavefront aberrations. On the image plane these are manifested in varying degrees of resolution and contrast degradation across the field-of-view. In pupil adaptive optics (AO), where a wavefront modulator is accommodated at the objective’s pupil-plane, a single correction profile is applied for all fields, and the correctable field-of-view (FoV) is limited by the isoplanatic patch. In the alternative configuration of conjugate adaptive optics, where the corrective element and the dominant aberrating layer within the sample sit at optically conjugate planes, effective correction across the entire FoV is possible in principle. This configuration, however, is relatively difficult to implement with deformable mirrors for wavefront modulation, since several folded optical paths have to be constructed. In this work, we present the design and evaluation of a completely in-line conjugate-AO system based on a refractive wavefront modulator (DELTA7 from Phaseform GmbH, Germany). We evaluate the performance of the system on a Zeiss Axiovert inverted microscope by imaging fluorescent beads through custom-fabricated phase plates.

The fluorescence Optofluidic imaging have recently transformed conventional optical investigations of biological systems by exploiting the synergic combination of recent microscopy and integrated chips technologies. The implementation of light-sheet microscopy on Optofluidic imaging has further pushed the systemic capabilities towards fast and 3D-volumetric imaging. Here, we propose a new compact open-top light-sheet microscope to extend the study of 3D-fluorescence Optofluidic imaging. The proposed portable architecture of the system and customized reconstruction algorithms enable providing high throughput and volumetric reconstruction capability with an isometric single-cell resolution. The experimental results demonstrate the applicability of open-top light-sheet configuration to single-cell imaging in a flow. We anticipate the system to offer promising applications in clinical settings where the portability and 3D-volumetric analysis are required for fast diagnostics.

We have developed wFLFM to further enhance the resolution by taking the advantage of both wide-field image and FLFM. We proposed the principle of wFLFM and verified the approach both numerically using simulated rings, and experimentally using phantom samples on the wFLFM system we constructed as a prototype. The principle of wFLFM requires only additional wild-field image at the aperture plane, making it readily compatible with various extended depth of focus wide-field modules, and the validating results showed a 2-3× enhancement in lateral resolution without cost in axial information, thereby advancing wFLFM as a promising high-resolution 3D imaging method.

Scanning Laser Optical Tomography (SLOT) is a three-dimensional imaging technique usable on a micro- to mesoscale. The technique is equivalent to computer tomography, using a laser instead of x-rays. With this technology, a set of twodimensional images is acquired at different angles and subsequently processed into volume information by reconstruction algorithms. Different contrast mechanisms can be used, depending on the application. Up until now transmission, scattering, fluorescence and second harmonic generation have been established for SLOT imaging. All of these contrast mechanisms are coupled to a specific narrow bandwidth, which is determined before the acquisition starts and is dependent on the setup and application. In order to collect true hyperspectral information, a spectrometer and a broadband light source have been integrated. This way, the amount of information is increased with each measurable wavelength, leading to various improvements of the SLOT technology. The entire transmission and absorption spectra of three-dimensional samples can now be measured and reconstructed. Here, we present the current state of development of the hyperspectral SLOT. This includes the technical construction of the hardware setup and the development of the software integration. Many challenges need to be overcome when implementing spectroscopy in a tomographic setup. Solutions to specific problems, such as decreased resolution and focal shift, will be presented. Finally, we will show the first results of hyperspectral SLOT imaging.

The triple-negative breast cancer (tnbc) is an aggressive subtype linked to a poor outcome of established breast cancer therapy. Increasing evidence points to the role of the tumor’s extracellular matrix (ECM) as a determinant of its aggressiveness as well as the effectiveness of chemical therapeutics. Three-dimensional imaging techniques can be used to unravel ECM architecture. Label-free contrast mechanisms such as second harmonic generation (SHG) avoid falsification and artifacts introduced by the labeling process. Here, we present the complementary use of two-photon excitation microscopy (TPEF) and Scanning Laser Optical Tomography (SLOT) for the investigation and quantification of tumor ECM. Both methods were used to capture fluorescence from antibody-labeled samples as well as the SHG signal from collagen strands in the ECM. SLOT generally allows for the investigation of larger samples of several mm up to a few cm in size. This work shows the capabilities of the tomographic setup compared to established TPEF, and demonstrates their combined use to maximize the information content of the acquired data. The obtained images served as a basis for ECM quantification. 3D-analysis allowed for determination of length, straightness and orientation of the collagen fibers based on fluorescence imaging as well as SHG imaging. The resulting coordinates might be used for synthetic reconstruction of a patient-specific tumor matrix, serving as a scaffold for pre-clinical therapeutic testing. Collagen imaging and quantification as presented here can therefore be employed for both basic and clinical research, paving the way for patient-specific cancer therapy.

Chondrocyte viability is an important measure to consider when assessing cartilage health. Dye-based cell viability assays are not suitable for in vivo or long-term studies. We have introduced a non-labeling viability assay based on the assessment of high-resolution images of cells and collagen structure using two-photon stimulated autofluorescence and second harmonic generation microscopy. By either the visual or quantitative assessment, we were able to differentiate living from dead chondrocytes in those images. However, both techniques require human participation and have limited throughputs. Throughput can be increased by using methods for automated cell-based image processing. Due to the poor image contrast, traditional image processing methods are ineffective on autofluorescence images produced by nonlinear microscopes. In this work, we examined chondrocyte segmentation and classification using Mask R-CNN, a deep learning approach to implement automated viability analysis. It has been demonstrated an 85% accuracy in chondrocyte viability assessment with proper training. This study demonstrates that automated and highly accurate image analysis is achievable with the use of deep learning methods. This image processing approach can be helpful to other imaging applications in clinical medicine and biological research.

Multi-photon microscopy (MPM) is a useful biomedical imaging tool due, in part, to its capabilities of probing tissue biomarkers at high resolution and with depth-resolved capabilities. Automated MPM tile scanning allows for whole-slide image acquisition but suffers from tile-stitching artifacts that prevent accurate quantitative data analysis. We have investigated a variety of post-processing artifact correction methods using ImageJ macros and custom Python/ MATLAB code and present a quantitative and qualitative comparison of these methods using whole-slide MPM autofluorescence images of human duodenal tissue. Image quality is assessed via evaluation of artifact removal compared to the calculated mean square error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM) of the processed image and its raw counterpart. Consideration of both quantitative and qualitative results suggest a combination of flat-field based correction and frequency filtering processing steps provide improved artifact correction when compared to each method used independently to correct for tiling artifacts of tile-scan MPM images.