Surgical resection is the most effective treatment strategy for solid tumors, but complete removal of the tumor is critical for post-surgical recovery/long-term survival and is dependent on correct identification of the tumor margin and accurate excision of microscopic residual tumor in the surgical field. Fluorescence image guided surgery is an emerging technique that has shown promise for intraoperative location of tumors and tumor margins. Current versions of such intraoperative fluorescence imaging, however, are generally limited to 2D near-surface images, i.e., without information about tumor depth. Here we present an intraoperative fluorescence imaging system for 3D volumetric imaging of tumors; the system uses spatial frequency domain diffuse optical tomography with an analytic inversion reconstruction method. The new instrument can derive depth-sensitive 3D tumor images at depths up to 1 cm, and it employs compact epi-imaging and illumination suitable for the operating room, with quasi-real-time image reconstruction for surgical visualization. We present experimental results with FDA-approved Indocynanine Green using an extensive array of tissue phantoms and in a pilot in-vivo study.

Scars can be debilitating and cause serious functional limitations, significantly reduced physical function and loss of ability to perform normal daily activities. Scar formation is not fully understood and the treatment options have been hampered by the lack of an objective diagnostic tool to assess scars. Presently, assessment of hypertrophic scars has been based on subjective clinician rankings using a four-parameter scale called the Vancouver Scar Scale (VSS) or the Patient Observer Scar Assessment Scale (POSAS) but no objective, standardized tool for quantifying scar severity is available, despite known inadequacies of the subjective scales. We have developed a hand-held multi modal system consisting of a combined Spatial Frequency Domain Imager (SFDI) used for the assessment of tissue molecular components and a polarimeter for structural measurements. The SFDI capability is provided by an Arduino board controlled spectrally and polarimetric diverse Light Emitting Diodes (LED) ring illuminator. For SFDI imagery, the LEDs are combined with sinusoidal patterns. A single pattern snapshot SFDI approach is used to observe and quantify the biological components in the scar tissue including: oxygenated and de oxygenated hemoglobin, water, and melanin. The SFDI system is integrated with a reduced Mueller Matrix polarimetric system, whose illumination is also included in the LED’s ring, and providing for the assessment of collagen orientation through Mueller Matrix decomposition. The design of the system and experimental work on phantoms will be presented.

In spatial frequency domain imaging (SFDI), a spatially modulated intensity pattern is projected on to tissue, with the demodulated reflectance having more superficial sensitivity with increasing spatial modulation frequency. With sub-diffusive SFDI, very high (>0.5 mm-1) spatial modulation frequencies are projected yielding sensitivity to the directionality of light scattering with only few scattering events occurring and sub-millimeter penetration depth and spatial resolution. This technique has been validated in a series of phantom experiments, where fractal distributions of polystyrene spheres were imaged, and through a model based inversion, the size scale distribution versus overall density of these particles could be separated and visualized in spatially resolved maps. With sensitivity to localized light scattering over a wide field of view (11 cm x 14 cm), this technique is being translated for the application of intraoperative breast tumor margin assessment. To test sensitivity to changes in human breast tissue morphology, a cohort of over 30 freshly excised human breast tissue specimens, including adipose, fibroglandular, fibroadenoma, and invasive carcinoma, have been imaged and co-registered to whole specimen histology. Statistical analysis of the distributions of both textual raw reflectance parameters and model based optical properties for each type of tissue will be presented. Furthermore, classification algorithm development and analysis to predicted likelihood of cancer on the surface of the tissue will also be presented. Reflectance maps, optical property maps, and probability likelihood maps of spatially heterogeneous samples with multiple tissue types will also be shown.

Fluorescence Lifetime Förster Resonance Energy Transfer (FLIM-FRET) is a unique non-invasive imaging platform to monitor and quantify in vivo target engagement in pre-clinical studies. FLIM FRET is a valuable tool in targeted drug delivery due to its nanoscale-range molecular resolution that detects near-infrared labeled ligand binding to dimerized receptors followed by their uptake into cancer cells in vivo. Various imaging platforms, including PET, lack the ability to directly discriminate between unbound and internalized ligands. Since transferrin receptor (TfR) level is significantly elevated in cancer cells compared to non-cancerous cells, transferrin (Tf) has been successfully used in molecular imaging and targeted anti-cancer drug delivery. The dimeric nature of TfR allows for the quantification of Tf internalization into cancer cells by measuring FLIM FRET between receptor-bound Tf donor and acceptor NIR fluorophore pairs, based on the reduction of donor fluorophore lifetime in live mice. We analyzed tumor morphology, the level of expression of TfR, estrogen receptor (ER) and Tf accumulation in human breast cancer tumor xenografts. We found a remarkable heterogeneity of breast cancer tumors regarding their size, cell density, TfR and ER expression and Tf uptake. The results of this study confirm a strong correlation between in vivo NIR FLIM FRET and ex vivo evaluation of Tf uptake into tumor tissues, thus validating FD% as a robust measure of the target engagement of TfR-Tf in tumor cells in vivo.

The surgeon’s limited ability to accurately delineate the tumor margin during surgical interventions is one key challenge in clinical management of cancer. New methods for guiding tumor resection decisions are needed. Numerous studies have shown that tissue autofluorescence properties have the potential to asses biochemical features associates with distinct pathologies in tissue and to distinguish various cancers from normal tissues. However, despite these promising reports, autofluorescence techniques were sparsely adopted in clinical settings. Moreover, when adopted they were primarily used for pre-operative diagnosis rather than guiding interventions. To address this need, we have researched and engineered instrumentation that utilizes label-free fluorescence lifetime contrast to characterize tissue biochemical features in vivo in patients and methodologies conducive to real-time (few seconds) diagnosis of tissue pathologies during surgical procedures. This presentation overviews clinically-compatible multispectral fluorescence lifetime imaging techniques developed in our laboratory and their ability to operate as stand-alone tools, integrated in a biopsy needle and in conjunction with the da Vinci surgical robot. We present pre-clinical and clinical studies in patients that demonstrate the potential of these techniques for intraoperative assessment of brain tumors and head and neck cancer. Current results demonstrate that intrinsic fluorescence signals can provide useful contrast for delineation distinct types of tissues including tumors intraoperatively. Challenges and solutions in the clinical implementation of these techniques are discussed.

Photoluminescence techniques are amongst the most widely used tools in the life sciences, with new and exciting applications in medical diagnostics and molecular imaging continuously emerging. Advantages include their comparative ease of use, unique sensitivity, non-invasive character, and potential for multiplexing, remote sensing, and miniaturization. General drawbacks are, however, signals, that contain unwanted wavelength- and polarization contributions from instrument-dependent effects, which are also time-dependent due to aging of instrument-components, and difficulties to measure absolute fluorescence intensities [1]. Moreover, scattering systems require special measurement geometries [2] and the interest in new optical reporters with emission > 1000 nm strategies for reliable measurements in the second diagnostic for the comparison of material performance and the rational design of new fluorophores with improved properties [3]. Here, we present strategies to versatile method-adaptable liquid and solid fluorescence standards for different fluorescence parameters including traceable instrument calibration procedures and the design of integrating sphere setups for the absolute measurement of emission spectra and quantum yields in the wavelength region of 350 to 1600 nm [4,5]. Examples are multi-emitter glasses, spectral fluorescence standards, and quantum yield standards for the UV/vis/NIR.

Fluorescent tissue phantoms are useful constructs in tracking the daily performance of a fluorescence imaging system. However, fluorescence imaging systems vary according to intended use, such as with an endoscope, or a camera with wide field optics. They also vary in terms of spectral bandwidth, or the sensor. We present a method on how the fluorescence measurement results of a calibrated tissue phantom from two different fluorescence imaging systems can be compared. This demonstrates how tissue phantoms, when calibrated with units of optical radiance, can be used beyond a single optical system.

In the development of new near-infrared (NIR) fluorescence dyes for image guided surgery, there is a need for new NIR sensitive camera systems that can easily be adjusted to specific wavelength ranges in contrast the present clinical systems that are only optimized for ICG. To test alternative camera systems, a setup was developed to mimic the fluorescence light in a tissue phantom to measure the sensitivity and resolution. Selected narrow band NIR LED’s were used to illuminate a 6mm diameter circular diffuse plate to create uniform intensity controllable light spot (μW-mW) as target/source for NIR camera’s. Layers of (artificial) tissue with controlled thickness could be placed on the spot to mimic a fluorescent ‘cancer’ embedded in tissue. This setup was used to compare a range of NIR sensitive consumer’s cameras for potential use in image guided surgery. The image of the spot obtained with the cameras was captured and analyzed using ImageJ software. Enhanced CCD night vision cameras were the most sensitive capable of showing intensities < 1 μW through 5 mm of tissue. However, there was no control over the automatic gain and hence noise level. NIR sensitive DSLR cameras proved relative less sensitive but could be fully manually controlled as to gain (ISO 25600) and exposure time and are therefore preferred for a clinical setting in combination with Wi-Fi remote control. The NIR fluorescence testing setup proved to be useful for camera testing and can be used for development and quality control of new NIR fluorescence guided surgery equipment.

5-ALA-PpIX fluorescence-guided brain tumour resection can increase the accuracy at which cancerous tissue is removed and thereby improve patient outcomes, as compared with standard white light imaging. Novel optical devices that aim to increase the specificity and sensitivity of PpIX detection are typically assessed by measurements in tissue-mimicking optical phantoms of which all optical properties are defined. Current existing optical phantoms specified for PpIX lack consistency in their optical properties, and stability with respect to photobleaching, thus yielding an unstable correspondence between PpIX concentration and the fluorescence intensity. In this study, we developed a set of aqueous-based phantoms with different compositions, using deionised water or PBS buffer as background medium, intralipid as scattering material, bovine haemoglobin as background absorber, and either PpIX dissolved in DMSO or a novel nanoparticle with similar absorption and emission spectrum to PpIX as the fluorophore. We investigated the phantom stability in terms of aggregation and photobleaching by comparing with different background medium and fluorophores, respectively. We characterised the fluorescence intensity of the fluorescent nanoparticle in different concentration of intralipid and haemoglobin and its time-dependent stability, as compared to the PpIX-induced fluorescence. We corroborated that the background medium was essential to prepare a stable aqueous phantom. The novel fluorescent nanoparticle used as surrogate fluorophore of PpIX presented an improved temporal stability and a reliable correspondence between concentration and emission intensity. We proposed an optimised phantom composition and recipe to produce reliable and repeatable phantom for validation of imaging device.

Near infrared fluorescence (NIRF) based image guided surgery aims to provide vital information to the surgeon in the operating room, such as locations of cancerous tissue that should be resected and healthy tissue that should to be preserved. Targeted molecular markers, such as tumor or nerve specific probes, are used in conjunctions with NIRF imaging and display systems to provide key information to the operator in real-time. One of the major hurdles for the wide adaptation of these imaging systems is the high cost to operate the instruments, large footprint and complexity of operating the systems. The emergence of wearable NIRF systems has addressed these shortcomings by minimizing the imaging and display systems’ footprint and reducing the operational cost. However, one of the major shortcomings for this technology is the replacement of the surgeon’s natural vision with an augmented reality view of the operating room. In this paper, we have addressed this major shortcoming by exploiting hologram technology from Microsoft HoloLens to present NIR information on a color image captured by the surgeon’s natural vision. NIR information is captured with a CMOS sensor with high quantum efficiency in the 800 nm wavelength together with a laser light illumination light source. The NIR image is converted to a hologram that is displayed on Microsoft HoloLens and is correctly co-registered with the operator’s natural eyesight.

Intraoperative resection of tumors currently relies upon the surgeon’s ability to visually locate and palpate tumor nodules. Undetected residual malignant tissue often results in the need for additional treatment or surgical intervention. The Solaris platform is a multispectral open-air fluorescence imaging system designed for translational fluorescence-guided surgery. Solaris supports video-rate imaging in four fixed fluorescence channels ranging from visible to near infrared, and a multispectral channel equipped with a liquid crystal tunable filter (LCTF) for multispectral image acquisition (520-620 nm). Identification of tumor margins using reagents emitting in the visible spectrum (400-650 nm), such as fluorescein isothiocyanate (FITC), present challenges considering the presence of auto-fluorescence from tissue and food in the gastrointestinal (GI) tract. To overcome this, Solaris acquires LCTF-based multispectral images, and by applying an automated spectral unmixing algorithm to the data, separates reagent fluorescence from tissue and food auto-fluorescence. The unmixing algorithm uses vertex component analysis to automatically extract the primary pure spectra, and resolves the reagent fluorescent signal using non-negative least squares. For validation, intraoperative in vivo studies were carried out in tumor-bearing rodents injected with FITC-dextran reagent that is primarily residing in malignant tissue 24 hours post injection. In the absence of unmixing, fluorescence from tumors is not distinguishable from that of surrounding tissue. Upon spectral unmixing, the FITC-labeled malignant regions become well defined and detectable. The results of these studies substantiate the multispectral power of Solaris in resolving FITC-based agent signal in deep tumor masses, under ambient and surgical light, and enhancing the ability to surgically resect them.

Plasmon resonance associated with nanoparticles of gold can enable photothermal ablation of tissues or controlled drug release with exquisite temporal and spatial control. These technologies may support many applications of precision medicine. However, clinical implementations of these technologies will require new methods of intraoperative imaging and guidance. Near-infrared laser surgery is a prime example that relies on improved image guidance. Here we set forth applications of augmented microscopy in guiding surgical procedures employing plasmon resonant gold-coated liposomes. Absorption of near-infrared laser light is the first step in activation of various diagnostic and therapeutic functions of these novel functional nanoparticles. Therefore, we demonstrate examples of near-infrared visualization of the laser beam and gold-coated liposomes. The augmented microscope proves to be a promising image guidance platform for a range of image-guided medical procedures.

Proteases are enzymes that play pathogenic roles in many common human diseases such as cancer, asthma, arthritis, atherosclerosis and infection by pathogens. Tools to dynamically monitor their activity can be used as diagnostic agents, as imaging contrast agents for intra-operative image guidance and for the identification of novel classes of protease-targeted drugs. I will describe our efforts to design and synthesize small molecule probes that produce a fluorescent signal upon binding to a protease target. We have identified probes that show tumor-specific retention, fast activation kinetics, and rapid systemic distribution making them useful for real-time fluorescence guided tumor resection and other diagnostic imaging applications.

Optical fluorescence-guided imaging is increasingly used to guide surgery and endoscopic procedures. Sprayable enzyme-activatable probes are particularly useful because of high target-to-background ratios that increase sensitivity for tiny cancer foci. However, green fluorescent activatable probes suffers from interference from autofluorescence found in biological tissue. Dynamic imaging followed by the kinetic analysis could be detected local enzyme activity and used to differentiate specific fluorescence arising from an activated probe in a tumor from autofluorescence in background tissues especially when low concentrations of the dye are applied to detect tiny cancer foci. Serial fluorescence imaging was performed using various concentrations of γ-glutamyl hydroxymethyl rhodamine green (gGlu-HMRG) which was sprayed on the peritoneal surface with tiny implants of SHIN3-dsRed ovarian cancer tumors. Temporal differences in signal between specific green fluorescence in cancer foci and non-specific autofluorescence in background tissue was measured and processed into three kinetic maps reflecting maximum fluorescence signal (MF), wash-in rate (WIR), and area under the curve (AUC), respectively. Especially at lower concentrations, kinetic maps derived from dynamic fluorescence imaging were clearly superior to unprocessed images for detection small cancer foci.

The development of drugs and devices to guide surgical resection of tumors in the United States requires the approval of the US Food and Drug Administration. Because these combine a drug and a device, the regulatory pathways can be confusing, particularly to academics or small companies. This paper discusses some of the issues and provides some guidance in this area.

Early detection of precursor lesions for colorectal cancer can greatly improve survival. Pre-neoplasia can appear flat with conventional white light endoscopy. Sessile serrated adenomas (SSA) are precursor lesions found primarily in the proximal colon and frequently appear flat and indistinct. We performed a clinical study of n=37 patients using a multimodal endoscopy with a FITC-labeled peptide specific for SSA. Lesions were imaged with white light, reflectance and fluorescence. White light images were acquired before the peptide was applied and were used to help localize regions of abnormal tissues rightly. Co-registered fluorescence and reflectance images were combined to get ratio images thus the distance was corrected. We calculated the target/background ratio (T/B ratio) to quantify the images and found 2.3-fold greater fluorescence intensity for SSA compared with normal tissues. We found the T/B ratio for SSA to be significantly greater than that for normal colonic mucosa with 89.47% sensitivity and 91.67% specificity at the threshold of 1.22. An ROC curve for SSA and normal mucosa was also plotted with area under curve (AUC) of 0.93. The result also shows that SSA and adenoma are statistically significant and can be identified with 78.95% sensitivity and 90.48% specificity at the threshold of 1.66. An ROC curve was plotted with AUC of 0.88. Therefore, our result shows that the application of a multimodal endoscope with fluorescently labeled peptide can quantify images and works especially good for the detection of SSA which is a premalignant flat lesion conferring a high risk of subsequently leading to a colon cancer.

Emerging optical imaging technologies can be integrated in the operating room environment during minimally invasive and open urologic surgery, including oncologic surgery of the bladder, prostate, and kidney. These technologies include macroscopic fluorescence imaging that provides contrast enhancement between normal and diseased tissue and microscopic imaging that provides tissue characterization. Optical imaging technologies that have reached the clinical arena in urologic surgery are reviewed, including photodynamic diagnosis, near infrared fluorescence imaging, optical coherence tomography, and confocal laser endomicroscopy. Molecular imaging represents an exciting future arena in conjugating cancer-specific contrast agents to fluorophores to improve the specificity of disease detection. Ongoing efforts are underway to translate optimal targeting agents and imaging modalities, with the goal to improve cancer-specific and functional outcomes.

Sarcomas are cancers of the bones, muscles, nerves, and fat that require complete surgical removal for cure. The primary surgical goal therefore is to remove the tumor with a zone of normal, non-cancerous tissue surrounding the tumor, considered a ‘negative’ surgical margin. At present, surgeons rely on radiologic imaging and visual and tactile clues to gauge cancer depth and guide surgical excision. This can result in removal of too much or too little tissue, which can lead to unnecessary removal of vital structures or incomplete cancer removal, respectively. Both results can have negative effects on ultimate patient outcome, with positive margins reported in 23% of sarcoma surgeries. Near-infrared (NIR) fluorescence probes are molecules that when stimulated with specific, known frequencies of near-infrared light will emit light of another distinct frequency. NIR light penetrates human tissue reasonably well and therefore can be used to detect the presence and location of unseen structures labeled with NIR fluorescence probes through several centimeters of tissue. Intra-operative near-infrared (NIR) fluorescence probes have been effective for this purpose in brain tumor surgery and may be applicable to sarcoma surgery. Foundational research is needed to explore the potential of this affibody probe and perfusion probes to estimate margin thickness in sarcoma surgery. In this study we will determine if sarcoma labeling using NIR fluorescence probes is superior with perfusion probes or a novel affibody probe. We will also determine whether NIR fluorescence using perfusion probes or a novel affibody probe can be correlated accurately to margin thickness.

Fluorescence imaging is a novel intraoperative technique for guiding oncologic surgeons toward radical tumor resections. Application of various fluorescent agents in exploratory clinical trials has already yielded promising results. The field of fluorescence imaging must now move beyond the proof-of-concept phase toward clinical application and implementation. This shift encompasses several hurdles, including standardization, advanced-phase study endpoints, regulatory affairs and routine implementation, which need to be addressed by the community in close collaboration with all stakeholders. These challenges, and the possible actions to overcome them and promote and accelerate clinical implementation of fluorescence imaging, are summarized in this paper.

Optical imaging detecting Cerenkov-light is a promising approach for molecular imaging or radiation therapy, but it was not yet conducted for proton therapy because light was not thought to be produced with the energy ranges because they are lower than Cerenkov-light threshold. Contrary to this consensus, our research group found that luminescence was emitted from water during proton-beam irradiation. The luminescence images of water phantoms showed clear Bragg peak, and the measured proton ranges from the images were almost the same as those obtained with an ionization chamber. The luminescence was also observed for carbon-ion and low energy X-ray photons.

A tomographic reconstruction algorithm for Cherenkov-excited luminescence scanned imaging (CELSI) is proposed and demonstrated for the first time, to reconstruct distributions of luminescent source. Coupled continuous wave (CW) diffusion equations are used to model luminescent photon propagation in biological tissues. The CELSI reconstruction was achieved by minimizing the difference between measured and computed data based on Tikhonov regularization technique. The feasibility and effectiveness of the algorithm were tested with numerical simulations on noisy data. In addition, comparisons between conventional diffuse optical fluorescence tomography (DOFT) and CELSI were also performed. Contrast-detail analysis was also used to evaluate the imaging performance of CELSI.

Light emitted as the result of high-energy particle transport through biological tissues (Cerenkov radiation) can be exploited for noninvasive diagnostic imaging using high sensitivity scientific cameras. We have investigated the energy transfer potential of Cerenkov radiation, discovering a new phototherapeutic technique for treatment of localized and disseminated cancers. This technique, Cerenkov radiation-induced phototherapy (CRIT), like photodynamic therapy, requires the presence of both light and photosensitive agent together to induce cytotoxicity and effective cancer treatment. But unlike conventional phototherapy strategies in which tissue ablation or activation of photoactive molecules is limited to superficial structures, radiation-induced phototherapy enables phototherapy delivery to the tumor sites throughout the body. Titanium oxide nanoparticles, which produce cytotoxic reactive oxygen species upon irradiation with UV light, were targeted to tumor tissue by surface decoration with transferrin. Subsequent administration of tumor-avid radiotracer, 18-fluorodeoxyglucose (18FDG) provided localized UV light source via Cerenkov radiation. Treatment of tumor-bearing mice with the combination of Titanium nanoparticles and 18FDG resulted in effective reduction in tumor growth, while individual agents were not therapeutic. This new strategy in cancer therapy extends the reach of phototherapy beyond what was previously possible, with potential for treatment of cancer metastases and rescue from treatment resistance.

Optical and ionizing radiation are two physical ways in which we can probe the living world. Until recently, these forms of radiation were used in distinct imaging and therapeutic applications—radiation therapy, photodynamic therapy, X-ray imaging, and diffuse optical tomography, to name a few. It has now been recognized that physical phenomena in which ionizing radiation and light are inherently coupled may provide powerful new capabilities for imaging and treating diseases. This presentation will review the physics and applications of radioluminescence, with a particular focus on molecular imaging. One such method, X-ray luminescence computed tomography (XLCT), uses narrow kilovolt X-ray beams to stimulate optical emissions from biologically targeted radioluminescent nanoparticles, thus providing high-resolution images even deep in tissue. A different phenomenon, Cherenkov luminescence, can also be harnessed to localize radiopharmaceuticals in vivo, allowing surgeons to visualize the molecular status of the tissues they are resecting. Recent progress towards routine implantation of these methods will be reviewed and sources of endogenous radioluminescence signal will be discussed.

Surgeons operating laparoscopically often have to rely upon subjective visual cues for complete oncological control and avoiding tumor violation or iatrogenic injury to critical tissues. A laparoscopic imaging tool to allow assessment of tumor margin or identification of anatomical structures buried under the layer of tissue being dissected is desirable to probe tissue contrast at a few millimeters depth, visualize over the lateral view for resection guidance, have a non-microscopic field-of-view (FOV) adequate for rapid survey of the resection site, and form the image in real-time. Probing light diffusely propagated through tissue provides sub-surface sensitivity, but the image formation generally involves intense computation that may be costly to intraoperative time-frame. Projecting these modalities laparoscopically to sample subsurface tissue heterogeneity over a non-microscopic FOV for rapid site-survey has been challenging. We demonstrate a laparoscopic applicator probe and a method thereof for real-time en-face mapping of near-surface heterogeneity for potential use towards intraoperative margin assessment. The probe fits a 12mm port and houses at 128 copper-coated 750μm fibers that form radially alternating illumination (70 fibers) and detection (58 fibers) channels. By simultaneously illuminating the 70 source channels of the laparoscopic probe that is in contact with a scattering medium and concurrently measuring the light diffusely propagated to the 58 detector channels, the presence of near-surface optical heterogeneities can be resolved in an en-face 9.5mm field-of-view in real-time. Visualization of subsurface margin of strong attenuation contrast at a depth up to 3mm is demonstrated at a frame rate of 1.25Hz.