ABSTRACT
BACKGROUND/OBJECTIVES: Pseudo-vascular network formation in vitro is considered a key characteristic of vasculogenic mimicry. While many cancer cell lines form pseudo-vascular networks, little is known about the spatiotemporal dynamics of these formations. METHODS: Here, we present a framework for monitoring and characterising the dynamic formation and dissolution of pseudo-vascular networks in vitro. The framework combines time-resolved optical microscopy with open-source image analysis for network feature extraction and statistical modelling. The framework is demonstrated by comparing diverse cancer cell lines associated with vasculogenic mimicry, then in detecting response to drug compounds proposed to affect formation of vasculogenic mimics. Dynamic datasets collected were analysed morphometrically and a descriptive statistical analysis model was developed in order to measure stability and dissimilarity characteristics of the pseudo-vascular networks formed. RESULTS: Melanoma cells formed the most stable pseudo-vascular networks and were selected to evaluate the response of their pseudo-vascular networks to treatment with axitinib, brucine and tivantinib. Tivantinib has been found to inhibit the formation of the pseudo-vascular networks more effectively, even in dose an order of magnitude less than the two other agents. CONCLUSIONS: Our framework is shown to enable quantitative analysis of both the capacity for network formation, linked vasculogenic mimicry, as well as dynamic responses to treatment.
Subject(s)
Neovascularization, Pathologic , Humans , Neovascularization, Pathologic/drug therapy , Neovascularization, Pathologic/pathology , Cell Line, Tumor , Melanoma/pathology , Melanoma/blood supply , Melanoma/drug therapy , Axitinib/pharmacologyABSTRACT
There is an urgent need to improve conventional cancer-treatments by preventing detrimental side effects, cancer recurrence and metastases. Recent studies have shown that presence of senescent cells in tissues treated with chemo- or radiotherapy can be used to predict the effectiveness of cancer treatment. However, although the accumulation of senescent cells is one of the hallmarks of cancer, surprisingly little progress has been made in development of strategies for their detection in vivo. To address a lack of detection tools, we developed a biocompatible, injectable organic nanoprobe (NanoJagg), which is selectively taken up by senescent cells and accumulates in the lysosomes. The NanoJagg probe is obtained by self-assembly of indocyanine green (ICG) dimers using a scalable manufacturing process and characterized by a unique spectral signature suitable for both photoacoustic tomography (PAT) and fluorescence imaging. In vitro, ex vivo and in vivo studies all indicate that NanoJaggs are a clinically translatable probe for detection of senescence and their PAT signal makes them suitable for longitudinal monitoring of the senescence burden in solid tumors after chemotherapy or radiotherapy.
Subject(s)
Cellular Senescence , Indocyanine Green , Indocyanine Green/chemistry , Cellular Senescence/drug effects , Humans , Animals , Optical Imaging , Mice , Nanoparticles/chemistry , Fluorescent Dyes/chemistry , Photoacoustic Techniques/methodsABSTRACT
Multispectral imaging captures spatial information across a set of discrete spectral channels and is widely utilized across diverse applications such as remote sensing, industrial inspection, and biomedical imaging. Multispectral filter arrays (MSFAs) are filter mosaics integrated atop image sensors that facilitate cost-effective, compact, snapshot multispectral imaging. MSFAs are pre-configured based on application-where filter channels are selected corresponding to targeted absorption spectra-making the design of optimal MSFAs vital for a given application. Despite the availability of many design and optimization approaches for spectral channel selection and spatial arrangement, major limitations remain. There are few robust approaches for joint spectral-spatial optimization, techniques are typically only applicable to limited datasets and most critically, are not available for general use and improvement by the wider community. Here, we reconcile current MSFA design techniques and present Opti-MSFA: a Python-based open-access toolbox for the centralized design and optimization of MSFAs. Opti-MSFA incorporates established spectral-spatial optimization algorithms, such as gradient descent and simulated annealing, multispectral-RGB image reconstruction, and is applicable to user-defined input of spatial-spectral datasets or imagery. We demonstrate the utility of the toolbox by comparing against other published MSFAs using the standard hyperspectral datasets Samson and Jasper Ridge, and further show application on experimentally acquired fluorescence imaging data. In conjunction with end-user input and collaboration, we foresee the continued development of Opti-MSFA for the benefit of the wider research community.
ABSTRACT
BACKGROUND: Ductal carcinoma in situ (DCIS) is a non-invasive form of early breast cancer, with a poorly understood natural history of invasive transformation. Necrosis is a well-recognized adverse prognostic feature of DCIS, and non-invasive detection of its presence and spatial extent could provide information not obtainable by biopsy. We describe here imaging of the distribution and extent of comedo-type necrosis in a model of human DCIS using C2Am, an imaging agent that binds to the phosphatidylserine exposed by necrotic cells. METHODS: We used an established xenograft model of human DCIS that mimics the histopathological features of the disease. Planar near-infrared and optoacoustic imaging, using fluorescently labeled C2Am, were used to image non-invasively the presence and extent of lesion necrosis. RESULTS: C2Am showed specific and sensitive binding to necrotic areas in DCIS tissue, detectable both in vivo and ex vivo. The imaging signal generated in vivo using near-infrared (NIR) fluorescence imaging was up to 6-fold higher in DCIS lesions than in surrounding fat pad or skin tissue. There was a correlation between the C2Am NIR fluorescence (Pearson R = 0.783, P = 0.0125) and optoacoustic signals (R > 0.875, P < 0.022) in the DCIS lesions in vivo and the corresponding levels of cell death detected histologically. CONCLUSIONS: C2Am is a targeted multi-modal imaging agent that could complement current anatomical imaging methods for detecting DCIS. Imaging the presence and spatial extent of necrosis may give better prognostic information than that obtained by biopsy alone.
Subject(s)
Breast Neoplasms/diagnostic imaging , Breast Neoplasms/pathology , Carcinoma in Situ/diagnostic imaging , Carcinoma in Situ/pathology , Carcinoma, Ductal, Breast/diagnostic imaging , Carcinoma, Ductal, Breast/pathology , Multimodal Imaging , Animals , Cell Death , Cell Line, Tumor , Contrast Media , Disease Models, Animal , Early Detection of Cancer , Female , Humans , Immunohistochemistry , Mice , Molecular Imaging , Multimodal Imaging/methods , Multimodal Imaging/standards , Optical Imaging , Photoacoustic TechniquesABSTRACT
Photoacoustic imaging (PAI) is an emerging tool that bridges the traditional depth limits of ballistic optical imaging and the resolution limits of diffuse optical imaging. Using the acoustic waves generated in response to the absorption of pulsed laser light, it provides noninvasive images of absorbed optical energy density at depths of several centimeters with a resolution of â¼100 µm. This versatile and scalable imaging modality has now shown potential for molecular imaging, which enables visualization of biological processes with systemically introduced contrast agents. Understanding the relative merits of the vast range of contrast agents available, from small-molecule dyes to gold and carbon nanostructures to liposome encapsulations, is a considerable challenge. Here we critically review the physical, chemical and biochemical characteristics of the existing photoacoustic contrast agents, highlighting key applications and present challenges for molecular PAI.
Subject(s)
Contrast Media/chemistry , Diagnostic Imaging/methods , Photoacoustic Techniques/methods , Whole Body Imaging/methods , Animals , Humans , Nanoparticles/chemistryABSTRACT
Flexible optical fibres, used in conventional medical endoscopy and industrial inspection, scramble phase and polarisation information, restricting users to amplitude-only imaging. Here, we exploit the near-diagonality of the multi-core fibre (MCF) transmission matrix in a parallelised fibre characterisation architecture, enabling accurate imaging of quantitative phase (error <0.3 rad) and polarisation-resolved (errors <10%) properties. We first demonstrate accurate recovery of optical amplitude and phase in two polarisations through the MCF by measuring and inverting the transmission matrix, and then present a robust Bayesian inference approach to resolving 5 polarimetric properties of samples. Our method produces high-resolution (9.0±2.6µm amplitude, phase; 36.0±10.4µm polarimetric) full-field images at working distances up to 1mm over a field-of-view up to 750×750µm 2 using an MCF with potential for flexible operation. We demonstrate the potential of using quantitative phase for computational image focusing and polarisation-resolved properties in imaging birefringence.
ABSTRACT
Rapid cancer cell proliferation promotes the production of reducing equivalents, which counteract the effects of relatively high levels of reactive oxygen species. Reactive oxygen species levels increase in response to chemotherapy and cell death, whereas an increase in antioxidant capacity can confer resistance to chemotherapy and is associated with an aggressive tumor phenotype. The pentose phosphate pathway is a major site of NADPH production in the cell, which is used to maintain the main intracellular antioxidant, glutathione, in its reduced state. Previous studies have shown that the rate of hyperpolarized [1-13C]dehydroascorbic acid (DHA) reduction, which can be measured in vivo using non-invasive 13C magnetic resonance spectroscopic imaging, is increased in tumors and that this is correlated with the levels of reduced glutathione. We show here that the rate of hyperpolarized [1-13C]DHA reduction is increased in tumors that have been oxidatively prestressed by depleting the glutathione pool by buthionine sulfoximine treatment. This increase was associated with a corresponding increase in pentose phosphate pathway flux, assessed using 13C-labeled glucose, and an increase in glutaredoxin activity, which catalyzes the glutathione-dependent reduction of DHA. These results show that the rate of DHA reduction depends not only on the level of reduced glutathione, but also on the rate of NADPH production, contradicting the conclusions of some previous studies. Hyperpolarized [1-13C]DHA can be used, therefore, to assess the capacity of tumor cells to resist oxidative stress in vivo However, DHA administration resulted in transient respiratory arrest and cardiac depression, which may prevent translation to the clinic.
Subject(s)
Dehydroascorbic Acid/metabolism , NADP/metabolism , Neoplasms/metabolism , Oxidative Stress , Animals , Carbon Isotopes , Cell Line, Tumor , Humans , Isotope Labeling , Magnetic Resonance Spectroscopy , MiceABSTRACT
BACKGROUND: Optoacoustic tomography (OT) of breast tumour oxygenation is a promising new technique, currently in clinical trials, which may help to determine disease stage and therapeutic response. However, the ability of OT to distinguish breast tumours displaying different vascular characteristics has yet to be established. The aim of the study is to prove OT as a sensitive technique for differentiating breast tumour models with manifestly different vasculatures. METHODS: Multispectral OT (MSOT) was performed in oestrogen-dependent (MCF-7) and oestrogen-independent (MDA-MB-231) orthotopic breast cancer xenografts. Total haemoglobin (THb) and oxygen saturation (SO2MSOT) were calculated. Pathological and biochemical evaluation of the tumour vascular phenotype was performed for validation. RESULTS: MCF-7 tumours show SO2MSOT similar to healthy tissue in both rim and core, despite significantly lower THb in the core. MDA-MB-231 tumours show markedly lower SO2MSOT with a significant rim-core disparity. Ex vivo analysis revealed that MCF-7 tumours contain fewer blood vessels (CD31+) that are more mature (CD31+/aSMA+) than MDA-MB-231. MCF-7 presented higher levels of stromal VEGF and iNOS, with increased NO serum levels. The vasculogenic process observed in MCF-7 was consistent with angiogenesis, while MDA-MB-231 appeared to rely more on vascular mimicry. CONCLUSIONS: OT is sensitive to differences in the vascular phenotypes of our breast cancer models.
Subject(s)
Biological Mimicry/physiology , Mammary Neoplasms, Experimental/blood supply , Mammary Neoplasms, Experimental/diagnosis , Mammary Neoplasms, Experimental/pathology , Neovascularization, Pathologic/diagnosis , Photoacoustic Techniques/methods , Tomography/methods , Animals , Breast Neoplasms/blood supply , Breast Neoplasms/diagnosis , Breast Neoplasms/pathology , Cell Line, Tumor , Drug Monitoring/methods , Female , Humans , MCF-7 Cells , Mammary Neoplasms, Experimental/metabolism , Mice , Mice, Inbred BALB C , Mice, Nude , Neoplasm Staging , Neovascularization, Pathologic/pathology , Oxygen Consumption/physiology , Sensitivity and Specificity , Tumor Hypoxia/physiology , Xenograft Model Antitumor AssaysABSTRACT
Raman microspectroscopy provides chemo-selective image contrast, sub-micrometer resolution, and multiplexing capabilities. However, it suffers from weak signals resulting in image-acquisition times of up to several hours. Surface-enhanced Raman scattering (SERS) can dramatically enhance signals of molecules in close vicinity of metallic surfaces and overcome this limitation. Multimodal, SERS-active nanoparticles are usually labeled with Raman marker molecules, limiting SERS to the coating material. In order to realize multimodal imaging while acquiring the rich endogenous vibronic information of the specimen, a core-shell particle based on "Nanorice", where a spindle-shaped iron oxide core is encapsulated by a closed gold shell, is developed. An ultrathin layer of silica prevents agglomeration and unwanted chemical interaction with the specimen. This approach provides Raman signal enhancement due to plasmon resonance effects of the shell while the optical absorption in the near-infrared spectral region provides contrast in photoacoustic tomography. Finally, T2-relaxation of a magnetic resonance imaging (MRI) experiment is altered by taking advantage of the iron oxide core. The feasibility for Raman imaging is evaluated by nearfield simulations and experimental studies on the primate cell line COS1. MRI and photoacoustics are demonstrated in agarose phantoms illustrating the promising translational nature of this strategy for clinical applications in radiology.
Subject(s)
Contrast Media/chemistry , Dust , Magnetic Resonance Imaging/methods , Nanoparticles/chemistry , Photoacoustic Techniques/methods , Spectrum Analysis, Raman , Animals , COS Cells , Chlorocebus aethiops , Computer Simulation , Nanoparticles/ultrastructure , Phantoms, ImagingABSTRACT
BACKGROUND AND STUDY AIMS: Endoscopic surveillance for Barrett's esophagus (BE) is limited by long procedure times and sampling error. Near-infrared (NIR) fluorescence imaging minimizes tissue autofluorescence and optical scattering. We assessed the feasibility of a topically applied NIR dye-labeled lectin for the detection of early neoplasia in BE in an ex vivo setting. METHODS: Consecutive patients undergoing endoscopic mucosal resection (EMR) for BE-related early neoplasia were recruited. Freshly collected EMR specimens were sprayed at the bedside with fluorescent lectin and then imaged. Punch biopsies were collected from each EMR under NIR light guidance. We compared the fluorescence intensity from dysplastic and nondysplastic areas within EMRs and from punch biopsies with different histological grades. RESULTS: 29 EMR specimens were included from 17 patients. A significantly lower fluorescence was found for dysplastic regions across whole EMR specimens (Pâ<â0.001). We found a 41â% reduction in the fluorescence of dysplastic compared to nondysplastic punch biopsies (Pâ<â0.001), with a sensitivity and specificity for dysplasia detection of 80â% and 82.9â%, respectively. CONCLUSION: Lectin-based NIR imaging can differentiate dysplastic from nondysplastic Barrett's mucosa ex vivo.
Subject(s)
Barrett Esophagus/diagnostic imaging , Esophageal Neoplasms/diagnostic imaging , Esophagus/pathology , Lectins/analysis , Molecular Imaging/methods , Optical Imaging/methods , Aged , Aged, 80 and over , Barrett Esophagus/pathology , Barrett Esophagus/surgery , Biopsy , Endoscopic Mucosal Resection , Esophageal Neoplasms/pathology , Esophageal Neoplasms/surgery , Feasibility Studies , Female , Fluorescence , Humans , Hyperplasia/diagnostic imaging , Hyperplasia/pathology , Male , Middle Aged , Sensitivity and Specificity , Staining and LabelingABSTRACT
Optical fiber technology is found in a wide variety of applications to flexibly relay light between two points, enabling information transfer across long distances and allowing access to hard-to-reach areas. Large-core optical fibers and light guides find frequent use in illumination and spectroscopic applications, for example, endoscopy and high-resolution astronomical spectroscopy. Proper alignment is critical for maximizing throughput in optical fiber coupling systems; however, there currently are no formal approaches to tolerancing the alignment of a light-guide coupling system. Here, we propose a Fourier alignment sensitivity (FAS) algorithm to determine the optimal tolerances on the alignment of a light guide by computing the alignment sensitivity. The algorithm shows excellent agreement with both simulated and experimentally measured values and improves on the computation time of equivalent ray-tracing simulations by two orders of magnitude. We then apply FAS to tolerance and fabricate a coupling system, which is shown to meet specifications, thus validating FAS as a tolerancing technique. These results indicate that FAS is a flexible and rapid means to quantify the alignment sensitivity of a light guide, widely informing the design and tolerancing of coupling systems.
ABSTRACT
Raman spectroscopy, amplified by surface enhanced Raman scattering (SERS) nanoparticles, is a molecular imaging modality with ultra-high sensitivity and the unique ability to multiplex readouts from different molecular targets using a single wavelength of excitation. This approach holds exciting prospects for a range of applications in medicine, including identification and characterization of malignancy during endoscopy and intraoperative image guidance of surgical resection. The development of Raman molecular imaging with SERS nanoparticles is presently limited by long acquisition times, poor spatial resolution, small field of view, and difficulty in animal handling with existing Raman spectroscopy instruments. Our goal is to overcome these limitations by designing a bespoke instrument for Raman molecular imaging in small animals. Here, we present a unique and dedicated small-animal Raman imaging instrument that enables rapid, high-spatial resolution, spectroscopic imaging over a wide field of view (> 6 cm(2)), with simplified animal handling. Imaging of SERS nanoparticles in small animals demonstrated that this small animal Raman imaging system can detect multiplexed SERS signals in both superficial and deep tissue locations at least an order of magnitude faster than existing systems without compromising sensitivity.
Subject(s)
Spectrum Analysis, Raman/methods , Animals , Female , Mice , Mice, NudeABSTRACT
In this paper a novel single-pixel method for coherent imaging through an endoscopic fiber bundle is presented. The use of a single-pixel detector allows greater sensitivity over a wider range of wavelengths, which could have significant applications in endoscopic fluorescence microscopy. First, the principle of lensless focussing at the distal end of a coherent fiber bundle is simulated to examine the impact of pixelation at microscopic scales. Next, an experimental optical correlator system using spatial light modulators (SLMs) is presented. A simple contrast imaging method of characterizing and compensating phase aberrations introduced by fiber bundles is described. Experimental results are then presented showing that our phase compensation method enables characterization of the optical phase profile of individual fiberlets. After applying this correction, early results demonstrating the ability of the system to electronically adjust the focal plane at the distal end of the fiber bundle are presented. The structural similarity index (SSIM) between the simulated image and the experimental focus-adjusted image increases noticeably when the phase correction is applied and the retrieved image is visually recognizable. Strategies to improve image quality are discussed.
ABSTRACT
Acute kidney injury (AKI) is a common and important medical problem, affecting 10% of hospitalized patients, and it is associated with significant morbidity and mortality. The most frequent cause of AKI is acute tubular necrosis (ATN). Current imaging techniques and biomarkers do not allow ATN to be reliably differentiated from important differential diagnoses, such as acute glomerulonephritis (GN). We investigated whether (13)C magnetic resonance spectroscopic imaging (MRSI) might allow the noninvasive diagnosis of ATN. (13)C MRSI of hyperpolarized [1,4-(13)C(2)]fumarate and pyruvate was used in murine models of ATN and acute GN (NZM2410 mice with lupus nephritis). A significant increase in [1,4-(13)C(2)]malate signal was identified in the kidneys of mice with ATN early in the disease course before the onset of severe histological changes. No such increase in renal [1,4-(13)C(2)]malate was observed in mice with acute GN. The kidney [1-(13)C]pyruvate/[1-(13)C]lactate ratio showed substantial variability and was not significantly decreased in animals with ATN or increased in animals with GN. In conclusion, MRSI of hyperpolarized [1,4-(13)C(2)]fumarate allows the detection of early tubular necrosis and its distinction from glomerular inflammation in murine models. This technique may have the potential to identify a window of therapeutic opportunity in which emerging therapies might be applied to patients with ATN, reducing the need for acute dialysis with its attendant morbidity and cost.
Subject(s)
Fumarates , Kidney Tubular Necrosis, Acute/diagnosis , Magnetic Resonance Imaging/methods , Animals , Carbon Isotopes , Early Diagnosis , Folic Acid , Humans , Kidney/abnormalities , Kidney/pathology , Kidney/physiopathology , Kidney Tubular Necrosis, Acute/chemically induced , Kidney Tubular Necrosis, Acute/physiopathology , Kinetics , Lupus Nephritis/diagnosis , Lupus Nephritis/pathology , Malates , Mice , Mice, Inbred C57BL , Pyruvic AcidABSTRACT
PURPOSE: The acquisition of ever increasing volumes of high resolution magnetic resonance imaging (MRI) data has created an urgent need to develop automated and objective image analysis algorithms that can assist in determining tumor margins, diagnosing tumor stage, and detecting treatment response. METHODS: We have shown previously that Minkowski functionals, which are precise morphological and structural descriptors of image heterogeneity, can be used to enhance the detection, in T1 -weighted images, of a targeted Gd(3+) -chelate-based contrast agent for detecting tumor cell death. We have used Minkowski functionals here to characterize heterogeneity in T2 -weighted images acquired before and after drug treatment, and obtained without contrast agent administration. RESULTS: We show that Minkowski functionals can be used to characterize the changes in image heterogeneity that accompany treatment of tumors with a vascular disrupting agent, combretastatin A4-phosphate, and with a cytotoxic drug, etoposide. CONCLUSIONS: Parameterizing changes in the heterogeneity of T2 -weighted images can be used to detect early responses of tumors to drug treatment, even when there is no change in tumor size. The approach provides a quantitative and therefore objective assessment of treatment response that could be used with other types of MR image and also with other imaging modalities.
Subject(s)
Etoposide/therapeutic use , Image Interpretation, Computer-Assisted/methods , Lymphoma/drug therapy , Lymphoma/pathology , Magnetic Resonance Imaging/methods , Stilbenes/therapeutic use , Animals , Antineoplastic Agents/therapeutic use , Cell Line, Tumor , Female , Mice , Mice, Inbred C57BL , Neoplasm Staging , Prognosis , Reproducibility of Results , Sensitivity and Specificity , Treatment OutcomeABSTRACT
Point-scanning microscopy approaches are transforming super-resolution imaging. Despite achieving parallel high-speed imaging using multifocal techniques, efficient multicolor imaging methods with high-quality illumination are currently lacking. In this paper, we present for the first time Mechanical-scan-free multiColor Super-resolution Microscopy (MCoSM) with spot array illumination, which enables mechanical-scan-free super-resolution imaging with adjustable resolution and a good effective field-of-view based on spatial light modulators. Through 100-2,500 s super-resolution spot illumination with different effective fields of view for imaging, we demonstrate the adjustable capacity of MCoSM. MCoSM extends existing spectral imaging capabilities through a time-sharing process involving different color illumination with phase-shift scanning while retaining the spatial flexibility of super-resolution imaging with diffractive spot array illumination. To demonstrate the prospects of MCoSM, we perform four-color imaging of fluorescent beads at high resolution. MCoSM provides a versatile platform for studying molecular interactions in complex samples at the nanoscale level.
ABSTRACT
Accurate measurement of optical absorption coefficients from photoacoustic imaging (PAI) data would enable direct mapping of molecular concentrations, providing vital clinical insight. The ill-posed nature of the problem of absorption coefficient recovery has prohibited PAI from achieving this goal in living systems due to the domain gap between simulation and experiment. To bridge this gap, we introduce a collection of experimentally well-characterised imaging phantoms and their digital twins. This first-of-a-kind phantom data set enables supervised training of a U-Net on experimental data for pixel-wise estimation of absorption coefficients. We show that training on simulated data results in artefacts and biases in the estimates, reinforcing the existence of a domain gap between simulation and experiment. Training on experimentally acquired data, however, yielded more accurate and robust estimates of optical absorption coefficients. We compare the results to fluence correction with a Monte Carlo model from reference optical properties of the materials, which yields a quantification error of approximately 20%. Application of the trained U-Nets to a blood flow phantom demonstrated spectral biases when training on simulated data, while application to a mouse model highlighted the ability of both learning-based approaches to recover the depth-dependent loss of signal intensity. We demonstrate that training on experimental phantoms can restore the correlation of signal amplitudes measured in depth. While the absolute quantification error remains high and further improvements are needed, our results highlight the promise of deep learning to advance quantitative PAI.
Subject(s)
Photoacoustic Techniques , Animals , Mice , Phantoms, Imaging , Photoacoustic Techniques/methods , Diagnostic Imaging , Computer Simulation , Monte Carlo MethodABSTRACT
Mesoscopic photoacoustic imaging (PAI) enables label-free visualization of vascular networks in tissues with high contrast and resolution. Segmenting these networks from 3D PAI data and interpreting their physiological and pathological significance is crucial yet challenging due to the time-consuming and error-prone nature of current methods. Deep learning offers a potential solution; however, supervised analysis frameworks typically require human-annotated ground-truth labels. To address this, an unsupervised image-to-image translation deep learning model is introduced, the Vessel Segmentation Generative Adversarial Network (VAN-GAN). VAN-GAN integrates synthetic blood vessel networks that closely resemble real-life anatomy into its training process and learns to replicate the underlying physics of the PAI system in order to learn how to segment vasculature from 3D photoacoustic images. Applied to a diverse range of in silico, in vitro, and in vivo data, including patient-derived breast cancer xenograft models and 3D clinical angiograms, VAN-GAN demonstrates its capability to facilitate accurate and unbiased segmentation of 3D vascular networks. By leveraging synthetic data, VAN-GAN reduces the reliance on manual labeling, thus lowering the barrier to entry for high-quality blood vessel segmentation (F1 score: VAN-GAN vs. U-Net = 0.84 vs. 0.87) and enhancing preclinical and clinical research into vascular structure and function.
Subject(s)
Deep Learning , Imaging, Three-Dimensional , Photoacoustic Techniques , Photoacoustic Techniques/methods , Humans , Imaging, Three-Dimensional/methods , Animals , Mice , Microvessels/diagnostic imaging , Breast Neoplasms/diagnostic imagingABSTRACT
Significance: Photoacoustic imaging (PAI) is an emerging technology that holds high promise in a wide range of clinical applications, but standardized methods for system testing are lacking, impeding objective device performance evaluation, calibration, and inter-device comparisons. To address this shortfall, this tutorial offers readers structured guidance in developing tissue-mimicking phantoms for photoacoustic applications with potential extensions to certain acoustic and optical imaging applications. Aim: The tutorial review aims to summarize recommendations on phantom development for PAI applications to harmonize efforts in standardization and system calibration in the field. Approach: The International Photoacoustic Standardization Consortium has conducted a consensus exercise to define recommendations for the development of tissue-mimicking phantoms in PAI. Results: Recommendations on phantom development are summarized in seven defined steps, expanding from (1) general understanding of the imaging modality, definition of (2) relevant terminology and parameters and (3) phantom purposes, recommendation of (4) basic material properties, (5) material characterization methods, and (6) phantom design to (7) reproducibility efforts. Conclusions: The tutorial offers a comprehensive framework for the development of tissue-mimicking phantoms in PAI to streamline efforts in system testing and push forward the advancement and translation of the technology.