This PDF file contains the front matter associated with SPIE Proceedings Volume 11625, including the Title Page, Copyright information, and Table of Contents.

This Conference Plenary, “Near-infrared nerve-specific probes to guide surgery,” was recorded for the Photonics West 2021 Digital Forum

Combining coherent Raman scattering microscopy and deuterium labeling provides insight into metabolism in live cancer cells during cancer development and progression. The dynamic data is modeled with Michaelis-Menten-kinetics to quantify lipid synthesis and utilization in cancer cells. Changes in lipid levels are found to originate from de novo lipid synthesis using glucose as a source. In this work, we isolate fatty acid synthesis/consumption rates and elucidated effects of altered lipid metabolism in T47D breast cancer cells in response to estradiol stimulation and etomoxir treatment, dynamic processes that cannot be easily observed without the application of appropriate models.

In this presentation I will describe our recent work to extend the spectral content of spatial frequency domain imaging (SFDI) to the short wave infrared (SWIR, 900-1300 nm). I will describe the unique instrumentation and methods needed for SWIR-SFDI measurements, and discuss the advantages and tradeoffs of imaging in the SWIR compared to the VIS and NIR. I will also present several preclinical and clinical application examples of SWIR-SFDI, including monitoring of edema, blood lipids, and spatial mapping of lipids in tumors.

Fluorescence lifetime imaging (FLI) has become an increasingly popular method in molecular imaging as it provides unique insights into the biological processes. However, despite its popularity and profound impact, FLI is not a direct imaging modality and datasets need to be postprocessed to quantify fluorescence lifetime or lifetime-based parameters. Such technical implementations can be complex, computationally expensive, require high level of expertise as well as user inputs. Herein, we will report on the development and validation of DL models as fast and user-friendly image formation tools for FLI, including outputting the quantitative lifetime image from raw FLI measurements without iterative solvers and user input, performing FLI topography corrected by the tissue optical properties and performing end-to-end 3D optical reconstructions.

Analyzing Spatial Frequency Domain Images (SFDI) of tissue in the sub-diffuse domain can reveal optical properties (μs’, γ) of the tissue related to its microstructural composition and shows potential for use in image-guided cancer removal. However, the determination of sub-diffuse optical properties is currently too slow for real-time applications. Recent research has demonstrated the real-time determination of these properties from experimental measurements using machine learning models, but the γ range of these models falls short of the full spectrum of γ values seen in biological tissue, limited by the range of the simulated datasets used to train these models. The Gegenbauer Kernel has previously been employed in SFDI simulations and been show to allow for simulations across an expanded γ range. Models trained on these simulations have shown success in simulation. We present a novel method which translates γ into analogous parameters of the Gegenbauer Kernel and uses this kernel to simulate datasets over an expanded range of γ values. We train a machine learning model on these datasets and use it to render sub-diffuse optical property heat maps from experimental data of tissue-simulating phantoms and ex vivo skin surgical samples across a full range of values in real-time. We compare this method against the current non-linear fit method and show a significant increase in speed with comparable accuracy. These findings enable real-time rendering of sub-diffuse SFDI for potential use within an image-guided surgery system.

Herein, we report on a depth-resolved Macroscopic Fluorescence Lifetime Imaging (MFLI) analytic framework based around machine learning coupled with a computationally efficient Monte Carlo-based data simulation workflow for robust and user-friendly model training. Our Siamese convolutional neural network (CNN) takes both optical properties (OPs) and time-resolved fluorescence decays as input and reconstructs both lifetime maps and depth profiles simultaneously. We validate our approach using phantom embeddings in silico and experimentally using Spatial Frequency Domain Imaging (SFDI) for OP retrieval. To our knowledge, this is the first study reporting the augmentation of MFLI with wide-field SFDI for lifetime topography.

The lack of quantitative information in image guided surgery determines still nowadays an unmet clinical need, leading to subjective assessments and variable outcomes. In this framework, we present the design of an endoscopic imaging system and the application of deep learning algorithms for real-time quantitation of tissues optical properties. The instrument is based on deep learning-optimized 3D profile corrected “Single Snapshot imaging of Optical Properties”(3D-SSOP). A first benchtop prototype has been validated on tissue mimicking phantoms and is currently being integrated on a surgical robot for pre-clinical trials on small animals.

Clinical trials with novel fluorescence contrast agents for head and neck cancer are driving new applications for fluorescence-guided surgery. Two-dimensional fluorescence imaging systems, however, provide limited in vivo assessment capabilities to determine tumor invasion depth below the mucosal surface. Here, we investigate the use of spatial frequency domain imaging (SFDI) methods for sub-surface fluorescence in tissue-simulating oral cancer phantoms. A two-step profile-correction approach for SFDI is under development to account for the complex surface topography of the oral cavity. First, for structured-illumination estimation of the surface profile, we are evaluating gray code and phase shift profilometry methods in agar-based oral cavity phantoms to maximize resolution and minimize sensitivity to surface discontinuities. Second, for profile-correction of the diffuse reflectance, global lighting effects within the oral cavity – analogous to an integrating sphere – are modeled using a multi-bounce numerical model. Subsurface fluorescence imaging is enabled based on the variations in optical sampling depth that result from changes in spatial frequency. An analytical depth recovery approach is based on a numerical diffusion theory model for semi-infinite fluorescence slabs of variable thickness. Depth estimation is evaluated in an agar-based phantom with fluorescence inclusions of thicknesses 1-5.5 mm originating from the top surface (“iceberg model”). Future clinical studies are necessary to assess in vivo performance and intraoperative workflow.

The staging of solid cancers is critical to the planning of both primary treatment with surgery and adjuvant therapies like chemotherapy; however, staging is not always possible with preoperative information and may require intraoperative evaluation of sentinel lymph nodes to confirm or disaffirm the presence of metastasis. Challenges are presented by standard-of-care sentinel lymph node dissection which must be quick and accurate enough to guide the surgical strategy despite a workflow that stretches from the operating room to the pathology lab; however, a solution is posed by fluorescence-assisted sentinel lymph node dissection which uses fluorescent probes to communicate the location and/or status of sentinel lymph nodes, reducing the complexity of the surgery and/or eliminating the need for rapid pathology. In support of this emerging modality, we have constructed a snapshot hyperspectral imaging system with sensitivity from the far-red to the near-infrared that enables sentinel lymph node dissection with multiple near-infrared fluorophores. We have also developed a spectral unmixing routine for in vivo quantification of the readily available fluorophores indocyanine green and methylene blue that can be extended to emerging fluorophores that actively target tumor cells. Both the imaging system and the unmixing routine have been tested in a clinical setting where they have successfully discriminated two dyes exhibiting different distributions.

Fluorescence-guided surgery give the surgeons extra input to improve the outcome of tumor resection procedures. However, the analysis of fluorescence images is qualitative and subjective inputs such as the surgeon’s perception and experience are considered when assessing the tumor margins. Objective indicators are needed to assess accurately the amount of fluorophore within the tissues. We developed a multimodal imaging platform capable of widefield quantitative fluorescence imaging for the use in a clinical environment. By mapping the fluorophore concentration, we offer an objective input for distinguishing healthy from diseased tissue and determining the resections margins in the optimal way: removing cancerous tissue while preserving healthy tissue and vital structures.

Intraoperative nerve recognition is critical to avoid accidental transection. Fluorescence-guided surgery can aid in nerve identification. However, detection of weak nerve-specific fluorescence signal is susceptible to the interference from high-background bright lights. We present a time-gated imager designed with ease-of-use and cost effectiveness in mind. Using this technology, we demonstrate successful rejection of room-light background signal to visualize the murine sciatic nerve.

Glioblastoma remains the deadliest type of brain tumor: half of patients do not live more than 16 months, even when treated with surgery, chemotherapy, and radiation. Tracking systems can help neurosurgeons precisely identify tumor on MRI images. Nevertheless, the tumor often regrows a few centimeters from where the original tumor was. This is because at the time of the first surgery, the actual tumor cells are already invading the healthy tissue around the tumor. These ‘invaders’ are difficult to cut out because even when looking through the surgical microscope, the tumor margins and normal brain tissue look very similar. The emerging utilization of fluorescing biomarkers (e.g., 5-ALA) sensitive to genetic downregulation present in cancer cells improves the detectability of marginal glioma, albeit requiring to switch to the surgical microscope excitation (blue light) mode and dim the operation room lights, imposing difficulties for neurosurgeons and staff. Here, we present a portable fluorescence-guided surgery optical imaging system integrated into the conventional surgical microscope to give neurosurgeons a better tool to predict which tissue is normal and which contains the start of tumor invasion without the need to switch to the excitation mode. The system operates under the microscope’s white light illumination using pulsed fluorophore excitation with gated acquisition and provides helpful tumor tissue fluorescing contrast. Tissue-mimicking phantom imaging confirmed protoporphyrin IX detection down to 0.1μg/mL concentration. Brain tissue imaging ex-vivo and pre-clinical intracranial tumor resection demonstrated the system’s capability to provide a typical operating environment with auxiliary or augmented visualization of PpIX possible.

Fluorescence imaging for surgical guidance is a proven modality that allows for visualization of fluorescent markers in numerous biological imaging applications, from cancer margins to tissue oxygenation. As the field continues to develop there is an urgent need for fluorescence-imaging standards that enable system characterization and performance monitoring. Here, we present an update on the proposed indocyanine green (ICG) matching imaging standard. The proposed standard is composed of three different tests: a varying concentration sensitivity test (1 nM-1000 nM), a tissue-equivalent-depth sensitivity test (0.5 mm-6 mm), and a 1951 USAF fluorescence resolution test. The new version of the standard incorporates a fully 3D printed design, which in which includes fluorescent, tissue-equivalent, and optically opaque material. Furthermore we provide photostability and NIST-traceable radiometric measurements.

Smartphones have been proposed for numerous biomedical imaging applications, many of which use fluorescence imaging techniques. While the most common applications are point-of-care testing and diagnostic imaging of external tissues, there are also recent reports on smartphone-based imaging for real-time treatment guidance (photodynamic therapy, endoscopy, and surgery). The most commonly cited reasons for incorporation of smartphones in biomedical imaging systems are cost and scalability; however, in some contexts, this may not be the paramount design constraint. As novel smartphone-based imaging systems continue to be developed for various clinical applications, it is important to establish guidelines for effective vs ineffective smartphone utilization in biomedical optics. In this presentation, we propose six guidelines for assessing smartphone utilization in biomedical imaging and apply them to emerging smartphone-based imaging systems within treatment guidance applications.

This Conference Presentation, “Cyanine fluorophore chemistry to enable dynamic and multicolor in vivo imaging,” was recorded for the Photonics West BiOS 2021 Digital Forum.

Iatrogenic nerve injury is a risk in many surgical procedures. We present our experience with a novel fluorescent agent to enhance the visualization of at-risk nerves during a prostatectomy in a canine model. Illuminare-1 is an intravenously administered myelin-binding fluorophore. After IRB approval and successfully visualizing target nerves in murine and porcine tests, we undertook a robotically-assisted prostatectomy on a dog, which has comparable genitourinary anatomy to a human. Dogs were positioned supine, anesthetized, the abdomen was insufflated, and the peri-prostatic and obturator nerves were exposed with a DaVinci SI Surgical System. A modified FDA-approved laparoscope with white and blue (370 – 425nm wave length) light settings was positioned via an assistant port to illuminate these nerves. A bolus of 1mg/kg of Illuminare-1 was administered intravenously with uninterrupted visualization of the presumed nerves. Fluorescent structures were resected for histological assessment. With 1mg/kg of Illuminare-1, nerves rapidly fluoresced under blue light, displaying a distinct hue. Fluorescence was seen within 90 seconds of administration and sustained for over three hours in the obturator and smaller periprostatic nerves. Tiny linear structures that were not initially seen under white light conditions were clearly identified as fluorescent tissue after injection with Illuminare-1. Seven fluorescent peri-prostatic structures were resected and all were histologically confirmed to be myelinated nerves. The cross-sectional nerve fiber diameters ranged from 64 – 247nm. Illuminare-1 enhanced the visualization of the neurovascular bundle in a dog. Phase-1 in-human trials with Illuminare-1 will follow to address the unmet need to reduce unintended surgical morbidity.

Accurate mapping of gastrointestinal stromal tumors (GIST) during surgery is difficult, which contributes to the suboptimal diagnosis and recurrence of cancers. To overcome this limitation, we developed a near-infrared (NIR) fluorescent nanoprobe for real-time navigation of GIST using a targeted strategy against the CD117 ligand stem cell factor (SCF). A zwitterionic NIR fluorophore conjugated to SCF showed specific binding to a xenograft mouse model of CD117-positive GIST-T1 with minimal nonspecific tissue signals. This promising intraoperative imaging strategy could be further explored for early diagnosis and follow-up of GIST prognosis before and after surgical resection.

Characterizing an administered drug’s pathway from initial systemic uptake, to targeted tissue accumulation, and the eventual excretion route is an important component of clinical translation. For mapping such pharmacokinetic behaviors in a biologically-relevant system, fluorescently-tagged drugs are commonly administered and examined in preclinical animal models. Broadband fluorescence cryo-imaging offers a high-resolution, whole-animal technique for recovering such fluorescently-tagged biodistributions, although agent-specificity remains a challenge due to unknown levels of heterogeneous tissue autofluorescence. Herein, we report on a new hyperspectral multichannel fluorescence cryo-imaging system and demonstrate higher agent-specificity and signal-sensitivity compared to conventional broadband fluorescence.

Concurrent administration of cancer therapeutics with tumor vasculature targeting treatment has been shown to improve overall survival in multiple human cancer types, as such combinations aim to destroy different compartments of tumors. Anti-angiogenesis therapeutics designed to inhibit tumor induced vessel sprouting have also been shown to re-model the tumor vasculature through a transient vessel normalization effect, which leads to improved perfusion of oxygen and drug in tumor. However, the effects that this normalized vasculature has on the availability of cancer receptor, such as EGFR, is unknown. Herein, we examined the use of MRI-PAFT to estimate cancer surface receptor availability in response to anti-angiogenesis therapy, using MRI-coupled paired agent fluorescence tomography. Bevacizumab treated tumors showed increase in RA compared to control tumors, but this was not statistically significant.

Nerve damage plagues surgical outcomes, significantly affecting post-surgical quality of life. Intraoperative nerve detection is difficult since neuroanatomy is varilable between patients, and nerves are typically protected deep within the tissue. Fluorescence-guided surgery (FGS) offers a potential means for enhanced intraoperative nerve identification and preservation. We have developed the first near infrared (NIR) nerve-specific fluorophores for use during FGS. Lead optimization has yielded water soluble derivatives with excellent safety and pharmacology parameters. Work is underway to plan and execute preclinical toxicity testing to enable first-in-human clincial trials.

Fluorescence-guided surgery (FGS) to aid in the precise visualization of vital nerve structures in real-time intraoperatively could greatly improve patient outcomes. We took a medicinal chemistry approach that facilitated the design of our first-in-class NIR nerve-binding small molecule fluorophore libraries with excitation and emission profiles compatible with the “700-” and “800-” nm fluorescence imaging channels in the clinical grade FGS systems. Molecular engineering of the lead candidates allowed for the development of water-soluble nerve-specific contrast agents with improved safety profile that has great potential for clinical translation in the near future.

Publisher’s Note: This conference presentation, originally published on 5 March 2021, was withdrawn on 15 March 2021 per author request.

Forty two patients with high energy open fractures were involved into the study to investigate whether an indocyanine green (ICG)-based dynamic contrast-enhanced fluorescence imaging (DCE-FI) can be used to objectively assess bone perfusion and guide surgical debridement. For each patient, fluorescence images were recorded after 0.1 mg/kg of ICG was administered intravenously. By utilizing a bone-specific kinetic model to the video sequences, the perfusion-related metrics were calculated. The results of this study shown that the quantitative ICG-based DEC-FI can accurately assess the human bone perfusion during the orthopedic surgery.

Fluorescent contrast agents targeted to cancer biomarkers are increasingly being explored for cancer detection, surgical guidance, and response monitoring. Efforts have been underway to topically apply such biomarker-targeted agents to freshly excised specimen for detecting cancer cell receptors on the surface as a method for intraoperative surgical margin assessment, including dual-probe staining methods introduce a second ‘non-specific’ optical agent as a control to help compensate for heterogeneous uptake and normalize the imaging field. Still, such specimen staining protocols introduce multifaceted complexity with unknown variables, such as tissue-specific diffusion, cell-specific binding and disassociation rates, and other factors, affecting the interpreted cancer-biomarker distribution across the specimen surface. The ability to recover three-dimensional dual-probe biodistributions throughout whole-specimens could offer a ground-truth validation method for examining topical staining uptake behaviors. Herein, we report on a novel method for characterizing dual-probe accumulation with 3D depth-profiles observed from a dual-probe fresh-specimen staining experiment.

Surgery and systemic therapy are the backbone treatment methods for solid tumors. However, incomplete surgical resection and poor response to systemic therapy have resulted in substantial cancer-related mortality and morbidity worldwide. Antibody-based fluorescence imaging holds great potential to enable precise surgical resection and improve our understanding of the mechanism driving resistance to systemic therapy in clinical tumors. In this talk, I will present our recent work on translating fluorescently labeled therapeutic antibodies for surgical navigation and for quantitating antibody delivery in solid tumors in first-in-human trials.