This PDF file contains the front matter associated with SPIE Proceedings Volume 12853, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

We develop six-dimensional single-molecule orientation-localization microscopy (SMOLM) to measure the 3D positions and 3D orientations simultaneously of single fluorophores. We show how careful optimization of phase and polarization modulation components can encode phase, polarization, and angular spectrum information from each fluorescence photon into a microscope’s dipole-spread function. We used the transient binding and blinking of Nile red (NR) to characterize the helical structure of fibrils formed by designed amphipathic peptides, KFE8^L and KFE8^D, and the pathological amyloid-beta peptide Aβ42. We also deployed merocyanine 540 to uncover the interfacial architectures of biomolecular condensates.

We propose the application of Compressed Sensing to Computational Scattered Light Imaging to decrease measurement time and data storage. Computational Scattered Light Imaging (ComSLI) determines three-dimensional fiber orientations and crossings in biomedical tissues like brain tissue. Currently, conventional ComSLI is time-consuming and generates large data. Compressed Sensing reconstructs signals with fewer samples than required by the Shannon-Nyquist theorem with minimal perceptual loss, significantly reducing the number of measurements. We introduce an optimized illumination strategy for ComSLI based on the Discrete Cosine Transform and validate it by reconstructing characteristic scattering patterns in vervet brain tissue, thereby demonstrating the feasibility of Compressed Sensing in ComSLI.

Fluorescence Lifetime Imaging (FLIM) is a powerful technique that measures the decay time of fluorophores present in tissue samples alluding to their constituent molecules. FLIM has gained popularity in biomedical imaging for applications such as detecting cancerous tumors, surgical guidance, etc. However, conventional FLIM systems are limited by a reduced number of spectral bands and long acquisition time. Moreover, the large footprint, complexity, and cost of the instrumentation make it difficult for clinical applications. In this paper, we demonstrate a reconstruction-based hyperspectral detector that can resolve decay time and intensities in broad spectral ranges while providing high sensitivity, high gain, and fast response time. The hyperspectral detector is comprised of an array of efficient, ultrafast avalanche photodetectors integrated with nanophotonic structures. We utilize different nanostructures in the detectors to modulate light-matter interactions in spectral channels. This allows us to computationally reconstruct the spectral profile of the incoming fluorescence spectrum without the need for additional filters or dispersive optics. Also, the nanophotonic structures enhance efficiency (by a factor of 2 to 10 over different wavelengths) while providing fast response time. An innovative detector design has been employed to reduce the breakdown of the avalanche photodetectors to -7.8V while maintaining high gain (~50) across the spectral range. Therefore, enabling low light detection with a high signal-to-noise ratio for FLIM applications. Added spectral channels would provide valuable information about tissue materials, morphology, and disease diagnosis. Such innovative hyperspectral sensors can now be integrated on-chip capable of miniaturizing the FLIM system and making it a commercially viable tool for clinical use. This technology has the potential to revolutionize the current FLIM system with improved detection capabilities opening doors for new horizons.

Cameras designed to operate at kHz frame rates suffer from higher noise levels due to increased readout rate. This can mean in scenarios where it is not possible to increase the illumination level, for example biomedical imaging applications, that objects may be lost in noise. The coherent gain effect inherent in an unbalanced holographic imaging system provides a solution, by amplifying the information in the signal beam with a bright reference beam. The resulting interference pattern can be detected above the noise floor of the camera, and intensity and phase images of the object reconstructed with off-axis holography techniques. This allows for a large improvement in the signal to noise ratio (SNR) of images of the object without having to increase the illumination level. We show that using this method it is possible to obtain images with a SNR ≥ 1 down to a detected intensity on the camera of ∼ 1 photon per pixel per frame.

Flow cytometry is a fundamental and powerful technique in the biomedical sciences which heavily depends on excitation lasers for imaging fluorescent markers to perform cell sorting and analysis. Modern multicolor flow cytometers incorporate several excitation lasers, typically 405 nm, 488 nm, 561 nm, and 640 nm allowing simultaneous cell characterization. Ultraviolet (UV) lasers, namely 355 nm, became indispensable flow cytometry excitation sources with the development of the Brilliant Ultraviolet (BUV) fluorochromes enabling high-dimensional analysis. Primary UV source for BUV excitation is represented by frequency tripled Nd:YVO₄ diode-pumped solid-state lasers (DPSSL) which constitute the single most expensive component of high-end flow cytometers. We demonstrate high-power UV generation from cascade intracavity frequency conversion based on the optically pumped vertically-external-cavity surface-emitting lasers (VECSELs). The active semiconductor gain medium is designed for operation around 1060 nm and is comprised of InGaAs quantum wells. The frequency conversion is implemented in a compact V-shaped cavity with the non-linear lithium triborate (LBO) crystals placed at the waist of the cavity mode in order to enhance conversion efficiency. This relatively novel and cost-effective semiconductor laser platform offers high-power, tunable, low-noise emission with excellent beam quality, important for flow cytometry applications. VECSEL platform can fill the spectral gaps where semiconductor diode laser technology is not available, namely green, and green-yellow spectral regions, and at the same time serving as a cost-effective replacement of DPSSL in the UV and DUV spectral bands.

Our high-speed multimodal imaging platform combines stimulated Raman Scattering (SRS), multiphoton fluorescence (MPF), and second harmonic generation (SHG) to explore metabolic activities of biomolecules in cells and tissues. Using the A-PoD algorithm, we achieve super-resolution imaging, while deuterium oxide probing enhances SRS imaging. The Correlation Coefficient Mapping (CoCoMap) technique helps understand coordination and regulation of metabolic activities. This tool enables early disease detection, prognosis, and therapeutic assessment, providing valuable insights into aging and metabolic diseases. Our approach revolutionizes metabolism visualization, advancing knowledge of biological processes and their implications in health and disease.

We have developed an automated single-molecule imaging system for live-cell analysis based on robotics and artificial intelligence-assisted microscopy. All significant procedures, i.e., searching for cells suitable for observation, detecting in-focus positions, and performing image acquisition and single-molecule tracking, are fully automated. Highly accurate, efficient, and reproducible single-molecule imaging experiments of receptor molecules on the plasma membrane of living cells can be performed. By narrowing the field of view to the size of a single cell, the excitation light is made to be uniform, allowing quantitative measurements of the diffusion and cluster formation of fluorescently labeled receptors. By developing a technology that automatically focuses on membrane receptors at high speed, high-throughput singlemolecule measurements that can observe about 400 fields of view per hour are achieved. We applied this technique to the single-molecule imaging analysis of epidermal growth factor receptors (EGFR). Changes in the lateral diffusion and cluster formation of EGFR on the membrane in response to EGF under various drug treatments, including tyrosine kinase inhibitors (TKIs), were clearly detected in individual cells. The spatiotemporal dynamics of EGFR upon its activation was characterized to determine several diffusion parameters and pharmacological parameters, including the number of diffusion states, diffusion coefficients, oligomer sizes, half-maximal effective concentration, and half-maximal inhibition concentration, showing a powerful potential for discovering novel therapeutics. Therefore, automated single-molecule imaging for systematic cell signaling analysis is feasible and is expected to extensively contribute to biological and pharmacological research.

The popularity of artificial intelligence (AI) and its applications has been steadily growing across various disciplines, including the field of science. In life sciences, AI has been increasingly applied for imaging flow cytometry towards automated management and classification of biological cell image data. However, when considering imaging flow cytometry for cell sorting, the need for employing AI over more facile, traditional image analysis methods are not as pronounced. This is primarily due to the time restrictions between image acquisition and sorting actuation that inevitably come with cell sorting. AI-enabled image-activated cell sorting (IACS) methods remain limited, and recent advancements in IACS still find success while relying on traditional feature gating strategies. Here, we compare the performance of feature gating, classical machine learning (ML), and deep learning (DL) in the differentiation of Saccharomyces cerevisiae mutant images taken by our intelligent IACS system to assess the advantages of AI for image classification in IACS. We show that while both classical ML and DL impart an increase in enrichment capability over feature gating, employing classical ML resulted in a smaller improvement at a larger cost of longer processing time than DL. We further performed IACS on mixed mutant populations and quantified target enrichment via downstream DNA sequencing to confirm the applicability of DL for the proposed study. Our findings validate the benefit and practicability of employing DL in IACS for microscopy-based genetic screening of S. cerevisiae, encouraging its inclusion for future advancements in this field.

We propose a method that combines high-speed Raman imaging with a machine learning technique, multi-armed bandit, to achieve rapid and accurate identification of samples under observation. First, our method dvides the field of view of the sample into small sections, and it returns either ’positive’ or ’negative’ based on whether the sections with high anomaly indices exceed a certain proportion. Moreover, the points to be measured are determined dynamically and automatically generating a series of optimal illumination patterns.

Drop-seq was first published in 2015, posing a revolution in single-cell RNA sequencing with an open-source, cost-effective, high-throughput analysis of gene transcripts. In this method, a highly diluted suspension of cells and barcoded beads is encapsulated in a water-in-oil droplet following double-Poisson statistics, aiming for a maximum encapsulation rate of one cell together with one bead, while keeping erroneous encapsulations of multiple cells or multiple beads together with one or more beads or cells, respectively, at a minimum. This decreases the efficiency and increases the cost per cell as more reagents are used. In this work, we present an approach to increase the number of desired encapsulation events, while decreasing erroneous encapsulations. Therefore, we use a special microfluidic analysis device in combination with a custom-made dielectrophoretic (DEP) sorting microchannel, enabling the detection and elimination of erroneous bead encapsulations, providing the opportunity to increase the concentration of beads to increase the number of desired encapsulation events (1 bead + 1 cell) considerably, while decreasing relevant erroneously encapsulations.

Intestinal anastomosis is a surgical procedure to reconnect two portions of the intestine and restore bowel continuity after removal of pathological elements. Anastomosis is required in nearly all the >330,000 bowel resections performed annually in the USA and England. Anastomotic leakage (AL) has been reported to happen in 2.6%-19% of cases depending on multiple factors such as the definition, location and type of anastomosis, as well as the cohort under investigation. Mortality rates due to AL vary between 10%-20%. AL occurs primarily when tissues are not healed after suture. Healing is affected by the restoration of local blood flow as well as matching tissue layers during suture. Ensuring this healing requires millimetric accuracy and consistency from surgeons despite the movement due to patient breathing and other factors, making anastomosis the most challenging step in gastrointestinal surgery. Current robotic surgery has advanced to enable laparoscopic surgery without human help. However, the potential of robotic surgery to lower AL rates is still hindered by the missing tissue identification due to the lack of tactile feedback by surgeons, which does not allow matching tissue layers during suture. Tissue identification can be performed by using diffuse reflectance spectroscopy with the aim of excluding fat and including only mucosa and muscle tissues in sutures. In this study, we have classify fat, mucosa and muscle tissues based on 3125 DRS measurements of freshly excised ex vivo specimens of 47 patients. DRS measurements were performed with fiber-optic probes of 630-μm source detector distance (SDD; probe 1) and 2500-μm SDD (probe 2) to measure tissue layers from ~0.5-1mm and from ~0.5-2 mm deep, respectively. By using probe 1 and 5-fold cross-validation of quadratic support vector machine (SVM) models, we obtained true positive rates of (97.9 ±1.8) % for fat tissues, (95.5 ±1.3) % for mucosa, and (95.5 ±1.1) % for muscle. Similarly for probe 2, we achieved true positive rates of (96.7 ±1.3) % for fat tissues, (95.7 ±1.0) % for mucosa, and (94.5 ±1.9) % for muscle.

±We are designing a real-time spectral imaging system using micro-LEDs and a high-speed micro-camera for potential endoscopic applications. Currently, gastrointestinal (GI) endoscopic imaging has largely been limited to white light imaging (WLI), while other endoscopic approaches have seen advancements in imaging techniques including fluorescence imaging, narrow-band imaging, and stereoscopic visualization. To further advance GI endoscopic imaging, we are working towards a high-speed spectral imaging system that can be implemented with a chip-on-tip design for a flexible endoscope. Hyperspectral imaging has potential applications in a variety of imaging procedures with its ability to discern spectral footprints of the imaging field in a series of two-dimensional images at different wavelengths. For investigating the feasibility of a real-time LED-based hyperspectral imaging system for endoscopic applications we designed and developed a large-scale prototype using through-hole LEDs, which is further analyzed as we design our future system. For high quality imaging, the LED array must be designed with the specific illumination patterns and intensities of each LED considered. We present our work in optimizing our current LED array through optical simulations performed in silico. Damped least squares and orthogonal descent algorithms are implemented to maximize irradiance power and improve illumination homogeneity at several working distances by adjusting radial distances of each LED from the camera. With our prototype LED-based hyperspectral imaging system, the simulation and optimization approach achieved an average increase of 8.36 ± 8.79% in irradiance and a 4.3% decrease in standard deviation at multiple working distances and field-of-views for each LED, as compared to the original design, leading to improved image quality and maintained acquisition speeds. This work highlights the value of in silico optical simulation and provides a framework for improved optical system design and will inform design decisions in future works.

Nanotechnology-driven advancements in analytical science have revolutionized disease detection, offering rapid and accurate diagnostic solutions. This study explores the potential of advanced spectroscopic techniques, including Surface Enhanced Raman Scattering (SERS) and Localized Surface Plasmon Resonance (LSPR), for the rapid detection of SARS CoV-2, the causative agent of COVID-19. Leveraging the principles of nanomaterial interactions with viral components, SERS and LSPR enable label-free detection with high sensitivity, specificity, and speed. SERS exploits enhanced Raman signals from virus-interacting nanoscale substrates, while LSPR relies on plasmon resonance shifts upon viral binding. We provide an in-depth analysis of both techniques' principles, instrumentation, and potential applications. Moreover, we discuss the integration of machine learning algorithms to quantify viral loads accurately. Results reveal the efficacy of SERS and LSPR in detecting SARS CoV-2 with sensitivity comparable to polymerase chain reaction (PCR) tests, offering real-time monitoring and potential point-of-care utility. The business prospects and commercialization potential of these technologies underscore their relevance for industries beyond diagnostics. With comprehensive insights into optimization strategies, challenges, and implications, this study contributes to the nascent field of Nanotechnology-based Analytics 4.0, driving the convergence of nanotechnology, spectroscopy, and artificial intelligence for innovative diagnostic solutions.