SignificanceCoronary heart disease has the highest rate of death and morbidity in the Western world. Atherosclerosis is an asymptomatic condition that is considered the primary cause of cardiovascular diseases. The accumulation of low-density lipoprotein triggers an inflammatory process in focal areas of arteries, which leads to the formation of plaques. Lipid-laden plaques containing a necrotic core may eventually rupture, causing heart attack and stroke. Lately, intravascular optical coherence tomography (IV-OCT) imaging has been used for plaque assessment. The interpretation of the IV-OCT images is performed visually, which is burdensome and requires highly trained physicians for accurate plaque identification.AimOur study aims to provide high throughput lipid-laden plaque identification that can assist in vivo imaging by offering faster screening and guided decision making during percutaneous coronary interventions.ApproachAn A-line-wise classification methodology based on time-series deep learning is presented to fulfill this aim. The classifier was trained and validated with a database consisting of IV-OCT images of 98 artery sections. A trained physician with expertise in the analysis of IV-OCT imaging provided the visual evaluation of the database that was used as ground truth for training and validation.ResultsThis method showed an accuracy, sensitivity, and specificity of 89.6%, 83.6%, and 91.1%, respectively. This deep learning methodology has the potential to increase the speed of lipid-laden plaques identification to provide a high throughput of more than 100 B-scans/s.ConclusionsThese encouraging results suggest that this method will allow for high throughput video-rate atherosclerotic plaque assessment through automated tissue characterization for in vivo imaging by providing faster screening to assist in guided decision making during percutaneous coronary interventions.
Coronary heart disease has the highest rate of death and morbidity in the Western world. Lipid-laden plaques containing a necrotic core may eventually rupture causing heart attack and stroke. Intravascular Optical Coherence Tomography (IV-OCT) imaging has been used for plaque assessment. However, the IV-OCT images are visually interpreted, which is burdensome and require highly trained physicians. This study aims to provide high throughput lipid-laden plaque identification that can assist in vivo imaging by offering faster screening and guided decision-making during percutaneous coronary interventions. An A-line wise classification methodology based on time-series deep learning is presented to fulfill this aim.
Ophthalmic OCT image-quality is highly variable and directly impacts clinical diagnosis of disease. Computational methods such as frame-averaging, filtering, deep-learning approaches are generally constrained by either extended imaging times when acquiring repeated-frames, over-smoothing and loss of features, or the need for extensive training sets. Self-fusion is a robust OCT image-enhancement method that overcomes these aforementioned limitations by averaging serial OCT frames weighted by their respective similarity. Here, we demonstrated video-rate self-fusion using a convolutional neural network. Our experimental results show a near doubling of OCT contrast-to-noise ratio at a frame-rate of ~22 fps when integrated with custom OCT acquisition software.
We present a deep-learning based approach for automated qualitative assessment of lesion volumes using OCT images to enable real-time assessment of injury severity and longitudinal tracking of tissues response to photodamage. The network has been trained to quantify photodamage between the outer plexiform layer (OPL) and retinal pigmented epithelium (RPE) accurately without the need for extensive image pre- and post-processing. Manually annotated OCT cross-sections were used as ground-truths to train a U-Net convolutional neural network. The network was designed and implemented in PyTorch based on the multi-scale U-Net architecture.
Multiphoton microscopy uses ultrafast nonlinear light-matter interactions to generate signal contrast from biological samples. The imaging of tissue from various organs plays an important role for a better understanding of cellular processes within their microenvironment and helps to reveal mechanisms of cellular changes in tissues during disease processes. Most tissue imaging studies by the pharmaceutical industry or by pathologists have typically been performed using harvested and sectioned tissue from organs to investigate drug toxicity or disease-related changes. However, immediately following biopsy, tissues begin to degrade due to cell necrosis and apoptosis, and substantial information is lost during the process. We demonstrate tissue degradation monitoring at different time points after tissue excision by using our label-free multimodal multiphoton imaging system which integrates SHG, TPEF, FLIM, and CARS in one platform. We examined whole organs and tissues harvested from mice, including kidney, liver, pancreas, and brain, and immersed each in several different media including saline, Euro-Collins solution, UW solution, HTK, and formalin. We collected time-lapse images from each sample and compared rates of cell degradation, tissue structure changes, and variations in optical properties including the intensities of NADH and FAD, the metabolic redox ratio, and FLIM of free/bound NADH. As a result, we quantified rates of degradation and metabolic changes associated with the preservation methods based on these label-free optical properties. Therefore, these results can be used as reference values for most ex vivo tissue research that relies on tissue and cell viability.
Pre-clinical toxicology is a statutory requirement of drug development and plays a significant role in reducing attrition in drug discovery. Histopathology and indirect methods such as measurement of toxicity-associated systemic markers in blood or urine samples are the state-of-the-art techniques for toxicity evaluation. Further improvements over these conventional techniques are needed to detect signs of drug-induced toxicity at earlier stages with higher sensitivity and specificity. Multiphoton nonlinear imaging techniques such as two-/three-photon microscopy (2PF/3PF), fluorescence lifetime imaging microscopy (FLIM), second/third harmonic generation (SHG/THG) and coherent anti-Stokes Raman scattering (CARS) microscopy can extract complimentary structural and metabolic information of the target tissue in a label-free manner. In this study, we investigated the capability of a multimodal multiphoton microscopy technique (2PF/3PF/SHG/THG/FLIM/CARS) for detecting both functional and structural changes associated with drug-induced toxicity. Cisplatin, a platinum-based chemotherapy drug, is a cytotoxic agent used to treat many types of cancers. Common side effects of Cisplatin include nephrotoxicity and gonadal dysfunction. We obtained multimodal optical images of organs such as kidney, liver, and testis harvested from mice treated with a single dose of Cisplatin (3mg/kg) by intraperitoneal injection. A control group was treated with 0.9% saline. Structural and metabolic biomarkers related to Cisplatin-induced toxicity were identified and characterized from these multimodal optical images obtained ex vivo. The preliminary results suggest that it may be possible to develop a novel platform for drug toxicity identification and assessment based on multimodal nonlinear optical imaging techniques.
A biological sample consists of a variety of complex biomolecules, and fluorescence microscopy enables visualization of specific molecules at the sub-cellular level. However, these fluorescence techniques require certain fluorescence dyes to label the sample, and the fluorophores raise serious problems such as photo toxicity and photobleaching which could affect biological functionality in living systems. Advanced label-free optical imaging techniques based on nonlinear optical phenomena overcome these limitations of fluorescence microscopy. We have developed a novel label-free multimodal multiphoton nonlinear optical imaging system based on a near-IR femtosecond laser with photonic crystal fiber and pulse shaper. This highly integrated system offers numerous label-free techniques including third harmonic generation, three-photon excited fluorescence, second harmonic generation, two-photon excited fluorescence, fluorescence lifetime imaging, and broadband coherent anti-Stokes Raman scattering microspectroscopy in one platform. All of the nonlinear signals are spectrally separated by dichroic filters and simultaneously measured by photomultiplier tubes. Moreover, this system includes phase-variance optical coherence tomography as well to enable vascular imaging. We have applied our system to investigate processes in numerous biological samples. Our imaging technique is highly integrated and time efficient to generate big data, offering an array of biomolecular information at one time without staining, three-dimensional sub-micron resolution with deeper penetration, and less photodamage. The big data output from this system is analyzed by multivariate analysis such as principal component analysis and hierarchical cluster analysis. Therefore, this novel technology and methodology will have a great impact on fast in vivo label-free biomedical imaging as a big data generator.
Atherosclerosis is a progressive asymptomatic disease that has the highest rate of death and morbidity in the United States. High macrophage infiltration and thin cap fibroatheromas are known to be the precursor lesions of plaque rupture. Lipid-laden macrophages called foam cells are formed by the uptake of lipids within the plaque. These foam cells eventually die forming a necrotic core. Ruptured plaques are characterized by a necrotic core with an overlying thin-ruptured cap highly infiltrated by macrophages. Imaging modalities capable of identifying macrophage clusters in atherosclerotic plaques could be used for plaque vulnerability assessment. In this study, Multispectral Fluorescence Lifetime Imaging (FLIM) is used to retrieve information of biochemical markers present in atherosclerotic tissue. Here, we present a computational methodology that makes use of FLIM-based biochemical plaque features in order to identify macrophage/foam cells in atherosclerotic plaques. In the proposed methodology, the FLIM lifetime map obtained from a spectral channel of 494 ± 20.5 nm provides information about the accumulation of macrophages, which produce long lifetimes (>6 ns). This methodology was validated against histopathological assessment (CD68 staining specific for macrophages) in terms of statistical correlation, a 10-fold cross validation (sensitivity = 88.45%; specificity= 91.21%), and receiver operating characteristic (ROC AUC = 0.91) analyses.
Acute cardiovascular events are still the leading cause of morbidity and mortality in the Western world. The rupture of a vulnerable atherosclerotic plaque is the most common cause of an acute cardiovascular event. Vulnerable plaques are characterized by presenting a necrotic core below a thin fibrous cap, and extensive infiltration of macrophages and foam cells. Thus, the degree of macrophage accumulation is an indicator in determining plaque progression and probability of rupture. This work presents a simple and fast image processing method for the identification of agglomerated macrophage/foam-cells regions in intravascular optical coherence tomography (IV-OCT) images that might be used for in-vivo assessment of plaque vulnerability. This method relies on the ratio of the values of either the normalized-intensity standard deviation or the entropy estimated over two axially adjacent regions of interest in IV-OCT images. This method is able of highlighting areas in the IV-OCT images where significant amount of macrophages is localized, and was applied to IV-OCT scans of 26 postmortem coronary segments and validated against histopathological assessment. The accuracy for detecting macrophage/foam-cell was as follows: the normalized standard deviation ratio approach showed an accuracy of 86% and a sensitivity and specificity of 85.8% and 86.1%, respectively; while the entropy ratio led to an accuracy of 86.9% and a sensitivity and specificity of 86.8% and 86.9%, respectively.
Toxicology of the male reproductive system has received increased interest in recent years partly fueled by the growing reports of falling sperm counts and rising reproductive disorders in the human population. Testicular toxicity (TT) in pharmaceutical development is a challenging issue due to the lack of simple and robust screening methods. Currently, histopathologic examination and hormonal evaluation are the commonly used methods to assess TT. Improved biomarker or screening platforms that would allow identification of TT at an earlier stage can have a significant impact on the safety evaluation of pharmaceutical candidates. We investigated the potential of label-free optical nonlinear imaging technologies such as fluorescence lifetime imaging microscopy (FLIM), multi-photon microscopy (MPM) and coherent anti-Stokes Raman scattering (CARS) microscopy to identify novel biomarkers for effective detection of TT. In this study, testicular damage was induced in rats by intraperitoneal injection with 3 mg/kg cisplatin, a chemotherapy drug. Multimodal optical images were obtained from the fixed, unstained testicular tissue sections of untreated and treated rats using a custom-built near-infrared multiphoton imaging system. Structural and biochemical parameters extracted from these images were compared between both groups to identify abnormal features associated with TT in the treated group. By analyzing the complimentary information obtained using these label-free optical imaging technologies, it may be possible to develop a novel platform for evaluation of TT in safety assessment of pharmaceuticals on reproduction and fertility, which reveal these changes at the molecular level and allow observation of these changes at an earlier time point than available today.
Approximately 29 million Americans have diabetes, and 86 million are living with prediabetes, increasing the risk of developing type 2 diabetes. Complications of wound healing in diabetic patients represent a significant health problem. Impaired diabetic wound healing is characterized by reduced collagen production and diminished angiogenesis. During the proliferative phase of wound healing, the injured tissue undergoes angiogenesis, re-epithelialization, and fibroplasia. Monitoring the development of new blood vessels, metabolic changes, and collagen deposition, is critical to elucidate the process of diabetic wound healing and to improve the development of therapeutic drugs. This study employs a custom-built multimodal microscope where Optical Coherence Tomography Angiography (OCTA) is used for studying neovascularization, Fluorescence Lifetime Imaging Microscopy (FLIM) for NADH/FAD assessment, Second Harmonic Generation (SHG) microscopy for analyzing collagen deposition, and Coherent anti-Stoke’s Raman Scattering (CARS) microscopy for visualizing water/lipid distribution, all together to non-invasively follow closure of a skin wound in healthy diabetic (db/db) mice treated with placebo and angiogenesis-promoting topical formulation (GlaxoSmithKline). The (db/db) mouse model presents hyperglycemia, obesity, and delayed wound healing that is pathologically similar to human type 2 diabetes mellitus. In this ongoing study, the animals are treated once daily for 14 days after wounding. Images of the wound and surrounding areas are taken at different time points for 28 days. In this experiment, the wound healing process is investigated to gain deeper understanding of the drug mechanism. The capability to non-invasively monitor wound healing mechanisms can become a valuable tool in development of new drug compounds for diabetic wound care.
In brain cancer surgery, it is critical to achieve extensive resection without compromising adjacent healthy, non-cancerous regions. Various technological advances have made major contributions in imaging, including intraoperative magnetic imaging (MRI) and computed tomography (CT). However, these technologies have pros and cons in providing quantitative, real-time and three-dimensional (3D) continuous guidance in brain cancer detection. Optical Coherence Tomography (OCT) is a non-invasive, label-free, cost-effective technique capable of imaging tissue in three dimensions and real time. The purpose of this study is to reliably and efficiently discriminate between non-cancer and cancer-infiltrated brain regions using OCT images. To this end, a mathematical model for quantitative evaluation known as the Blind End- Member and Abundances Extraction method (BEAE). This BEAE method is a constrained optimization technique which extracts spatial information from volumetric OCT images. Using this novel method, we are able to discriminate between cancerous and non-cancerous tissues and using logistic regression as a classifier for automatic brain tumor margin detection. Using this technique, we are able to achieve excellent performance using an extensive cross-validation of the training dataset (sensitivity 92.91% and specificity 98.15%) and again using an independent, blinded validation dataset (sensitivity 92.91% and specificity 86.36%). In summary, BEAE is well-suited to differentiate brain tissue which could support the guiding surgery process for tissue resection.
In brain cancer surgery, it is critical to achieve extensive resection without compromising adjacent healthy, noncancerous regions. Various technological advances have made major contributions in imaging, including intraoperative magnetic imaging (MRI) and computed tomography (CT). However, these technologies have pros and cons in providing quantitative, real-time and three-dimensional (3D) continuous guidance in brain cancer detection. Optical Coherence Tomography (OCT) is a non-invasive, label-free, cost-effective technique capable of imaging tissue in three dimensions and real time. The purpose of this study is to reliably and efficiently discriminate between non-cancer and cancerinfiltrated brain regions using OCT images. To this end, a mathematical model for quantitative evaluation known as the Blind End-Member and Abundances Extraction method (BEAE). This BEAE method is a constrained optimization technique which extracts spatial information from volumetric OCT images. Using this novel method, we are able to discriminate between cancerous and non-cancerous tissues and using logistic regression as a classifier for automatic brain tumor margin detection. Using this technique, we are able to achieve excellent performance using an extensive cross-validation of the training dataset (sensitivity 92.91% and specificity 98.15%) and again using an independent, blinded validation dataset (sensitivity 92.91% and specificity 86.36%). In summary, BEAE is well-suited to differentiate brain tissue which could support the guiding surgery process for tissue resection.
We present a wide-field fluorescence lifetime imaging (FLIM) system with optical sectioning by structured illumination microscopy (SIM). FLIM measurements were made using a time gated ICCD camera in conjunction with a pulsed nitrogen dye laser operating at 450 nm. Intensity images were acquired at multiple time delays from a trigger initiated by a laser pulse to create a wide-field FLIM image, which was then combined with three phase SIM to provide optical sectioning. Such a mechanism has the potential to increase the reliability and accuracy of the FLIM measurements by rejecting background intensity. SIM also provides the opportunity to create volumetric FLIM images with the incorporation of scanning mechanisms for the sample plane. We present multiple embodiments of such a system: one as a free space endoscope and the other as a fiber microendoscope enabled by the introduction of a fiber bundle. Finally, we demonstrate the efficacy of such an imaging system by imaging dyes embedded in a tissue phantom.
A novel computational method for plaque tissue characterization based on Intravascular Optical Coherence Tomography (IV-OCT) is presented. IV-OCT is becoming a powerful tool for the clinical evaluation of atherosclerotic plaques; however, it requires a trained expert for visual assessment and interpretation of the imaged plaques. Moreover, due to the inherit effect of speckle and the scattering attenuation of the optical scheme the direct interpretation of OCT images is limited. To overcome these difficulties, we propose to automatically identify the A-line profiles of the most significant plaque types (normal, fibrotic, or lipid-rich) and their respective abundance by using a probabilistic framework and blind alternated least squares to achieve the optimal decomposition. In this context, we present preliminary results of this novel probabilistic classification tool for intravascular OCT that relies on two steps. First, the B-scan is pre-processed to remove catheter artifacts, segment the lumen, select the region of interest (ROI), flatten the tissue surface, and reduce the speckle effect by a spatial entropy filter. Next, the resulting image is decomposed and its A-lines are classified by an automated strategy based on alternating-least-squares optimization. Our early results are encouraging and suggest that the proposed methodology can identify normal tissue, fibrotic and lipid-rich plaques from IV-OCT images.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.