Implementation of a portable diffraction phase microscope for digital histopathology.

Quantitative phase imaging (QPI) has a significant advantage in histopathology as it helps in differentiating biological tissue structures and cells without the need for staining. To make this capability more accessible, it is crucial to develop compact and portable systems. In this study, we introduce a portable diffraction phase microscopy (DPM) system that allows the acquisition of phase map images from various organs in mice using a low-NA objective lens. Our findings indicate that the cell and tissue structures observed in portable DPM images are similar to those seen in conventional histology microscope images. We confirmed that the developed system's performance is comparable to the benchtop DPM system. Additionally, we investigate the potential utility of digital histopathology by applying deep learning technology to create virtual staining of DPM images.

Multi-contrast digital histopathology of mouse organs using quantitative phase imaging and virtual staining.

Quantitative phase imaging (QPI) has emerged as a new digital histopathologic tool as it provides structural information of conventional slide without staining process. It is also capable of imaging biological tissue sections with sub-nanometer sensitivity and classifying them using light scattering properties. Here we extend its capability further by using optical scattering properties as imaging contrast in a wide-field QPI. In our first step towards validation, QPI images of 10 major organs of a wild-type mouse have been obtained followed by H&E-stained images of the corresponding tissue sections. Furthermore, we utilized deep learning model based on generative adversarial network (GAN) architecture for virtual staining of phase delay images to a H&E-equivalent brightfield (BF) image analogues. Using the structural similarity index, we demonstrate similarities between virtually stained and H&E histology images. Whereas the scattering-based maps look rather similar to QPI phase maps in the kidney, the brain images show significant improvement over QPI with clear demarcation of features across all regions. Since our technology provides not only structural information but also unique optical property maps, it could potentially become a fast and contrast-enriched histopathology technique.

In vivo electrochemical lipolysis of fat in a Yucatan pig model: A proof of concept study.

OBJECTIVES: Traditional fat contouring is now regularly performed using numerous office- based less invasive techniques. However, some limitations of these minimally invasive techniques include high cost or limited selectivity with performing localized contouring and reduction of fat. These shortcomings may potentially be addressed by electrochemical lipolysis (ECLL), a novel approach that involves the insertion of electrodes into tissue followed by application of a direct current (DC) electrical potential. This results in the hydrolysis of tissue water creating active species that lead to fat necrosis and apoptosis. ECLL can be accomplished using a simple voltage-driven system (V-ECLL) or a potential-driven feedback cell (P-ECLL) both leading to water electrolysis and the creation of acid and base in situ. The aim of this study is to determine the long-lasting effects of targeted ECLL in a Yucatan pig model. METHODS: A 5-year-old Yucatan pig was treated with both V-ECLL and P-ECLL in the subcutaneous fat layer using 80:20 platinum:iridium needle electrodes along an 8 cm length. Dosimetry parameters included 5 V V-ECLL for 5, 10, and 20 minutes, and -1.5 V P-ECLL, -2.5 V P-ECLL, -3.5 V P-ECLL for 5 minutes. The pig was assessed for changes in fat reduction over 3 months with digital photography and ultrasound. After euthanasia, tissue sections were harvested and gross pathology and histology were examined. RESULTS: V-ECLL and P-ECLL treatments led to visible fat reduction (12.1%-27.7% and 9.4%-40.8%, respectively) and contour changes across several parameters. An increased reduction of the superficial fat layer occurred with increased dosimetry parameters with an average charge transfer of 12.5, 24.3, and 47.5 C transferred for 5 V V-ECLL for 5, 10, and 20 minutes, respectively, and 2.0, 11.5, and 24.0 C for -1.5 V P-ECLL, -2.5 V P-ECLL, -3.5 V P-ECLL for 5 minutes, respectively. These dose-dependent changes were also evidenced by digital photography, gross pathology, ultrasound imaging, and histology. CONCLUSIONS: ECLL results in selective damage and long-lasting changes to the adipose layer in vivo. These changes are dose-dependent, thus allowing for more precise contouring of target areas. P-ECLL has greater efficiency and control of total charge transfer compared to V-ECLL, suggesting that a low-voltage potentiostat treatment can result in fat apoptosis equivalent to a high-voltage DC system.

Deep Learning-Based Glaucoma Screening Using Regional RNFL Thickness in Fundus Photography.

Since glaucoma is a progressive and irreversible optic neuropathy, accurate screening and/or early diagnosis is critical in preventing permanent vision loss. Recently, optical coherence tomography (OCT) has become an accurate diagnostic tool to observe and extract the thickness of the retinal nerve fiber layer (RNFL), which closely reflects the nerve damage caused by glaucoma. However, OCT is less accessible than fundus photography due to higher cost and expertise required for operation. Though widely used, fundus photography is effective for early glaucoma detection only when used by experts with extensive training. Here, we introduce a deep learning-based approach to predict the RNFL thickness around optic disc regions in fundus photography for glaucoma screening. The proposed deep learning model is based on a convolutional neural network (CNN) and utilizes images taken with fundus photography and with RNFL thickness measured with OCT for model training and validation. Using a dataset acquired from normal tension glaucoma (NTG) patients, the trained model can estimate RNFL thicknesses in 12 optic disc regions from fundus photos. Using intuitive thickness labels to identify localized damage of the optic nerve head and then estimating regional RNFL thicknesses from fundus images, we determine that screening for glaucoma could achieve 92% sensitivity and 86.9% specificity. Receiver operating characteristic (ROC) analysis results for specificity of 80% demonstrate that use of the localized mean over superior and inferior regions reaches 90.7% sensitivity, whereas 71.2% sensitivity is reached using the global RNFL thicknesses for specificity at 80%. This demonstrates that the new approach of using regional RNFL thicknesses in fundus images holds good promise as a potential screening technique for early stage of glaucoma.

Exploring feedback-controlled versus open-circuit electrochemical lipolysis in ex vivo and in vivo porcine fat: A feasibility study.

OBJECTIVES: Minimally invasive fat sculpting techniques are becoming more widespread with the development of office-based devices and therapies. Electrochemical lipolysis (ECLL) is a needle-based technology that uses direct current (DC) to electrolyze tissue water creating acid and base in situ. In turn, fat is saponified and adipocyte cell membrane lysis occurs. The electrolysis of water can be accomplished using a simple open-loop circuit (V-ECLL) or by incorporating a feedback control circuit using a potentiostat (P-ECLL). A potentiostat utilizes an operational amplifier with negative feedback to allow users to precisely control voltage at specific electrodes. To date, the variation between the two approaches has not been studied. The aim of this study was to assess current and charge transfer variation and lipolytic effect created by the two approaches in an in vivo porcine model. METHODS: Charge transfer measurements from ex vivo V-ECLL and P-ECLL treated porcine skin and fat were recorded at -1 V P-ECLL, -2 V P-ECLL, -3 V P-ECLL, and -5 V V-ECLL each for 5 min to guide dosimetry parameters for in vivo studies. In follow-up in vivo studies, a sedated female Yorkshire pig was treated with both V-ECLL and P-ECLL across the dorsal surface over a range of dosimetry parameters, including -1.5 V P-ECLL, -2.5 V P-ECLL, -3.5 V P-ECLL, and 5 V V-ECLL each treated for 5 min. Serial biopsies were performed at baseline before treatment, 1, 2, 7, 14, and 28 days after treatment. Tissue was examined using fluorescence microscopy and histology to compare the effects of the two ECLL approaches. RESULTS: Both V-ECLL and P-ECLL treatments induced in-vivo fat necrosis evident by adipocyte membrane lysis, adipocyte denuclearization, and an acute inflammatory response across a 28-day longitudinal study. However, -1.5 V P-ECLL produced a smaller spatial necrotic effect compared to 5 V V-ECLL. In addition, 5 V V-ECLL produced a comparable necrotic effect to that of -2.5 V and -3.5 V P-ECLL. CONCLUSIONS: V-ECLL and P-ECLL at the aforementioned dosimetry parameters both achieved fat necrosis by adipocyte membrane lysis and denuclearization. The -2.5 V and -3.5 V P-ECLL treatments created spatially similar fat necrotic effects when compared to the 5 V V-ECLL treatment. Quantitatively, total charge transfer between dosimetry parameters suggests that -2.5 V P-ECLL and 5 V V-ECLL produce comparable electrochemical reactions. Such findings suggest that a low-voltage closed-loop potentiostat-based system is capable of inducing fat necrosis to a similar extent compared to that of a higher voltage direct current system.

Quantitative Screening of Cervical Cancers for Low-Resource Settings: Pilot Study of Smartphone-Based Endoscopic Visual Inspection After Acetic Acid Using Machine Learning Techniques.

BACKGROUND: Approximately 90% of global cervical cancer (CC) is mostly found in low- and middle-income countries. In most cases, CC can be detected early through routine screening programs, including a cytology-based test. However, it is logistically difficult to offer this program in low-resource settings due to limited resources and infrastructure, and few trained experts. A visual inspection following the application of acetic acid (VIA) has been widely promoted and is routinely recommended as a viable form of CC screening in resource-constrained countries. Digital images of the cervix have been acquired during VIA procedure with better quality assurance and visualization, leading to higher diagnostic accuracy and reduction of the variability of detection rate. However, a colposcope is bulky, expensive, electricity-dependent, and needs routine maintenance, and to confirm the grade of abnormality through its images, a specialist must be present. Recently, smartphone-based imaging systems have made a significant impact on the practice of medicine by offering a cost-effective, rapid, and noninvasive method of evaluation. Furthermore, computer-aided analyses, including image processing-based methods and machine learning techniques, have also shown great potential for a high impact on medicinal evaluations. OBJECTIVE: In this study, we demonstrate a new quantitative CC screening technique and implement a machine learning algorithm for smartphone-based endoscopic VIA. We also evaluated the diagnostic performance and practicability of the approach based on the results compared to the gold standard and from physicians' interpretation. METHODS: A smartphone-based endoscope system was developed and applied to the VIA screening. A total of 20 patients were recruited for this study to evaluate the system. Overall, five were healthy, and 15 were patients who had shown a low to high grade of cervical intraepithelial neoplasia (CIN) from both colposcopy and cytology tests. Endoscopic VIA images were obtained before a loop electrosurgical excision procedure for patients with abnormal tissues, and their histology tissues were collected. Endoscopic VIA images were assessed by four expert physicians relative to the gold standard of histopathology. Also, VIA features were extracted from multiple steps of image processing techniques to find the differences between abnormal (CIN2+) and normal (≤CIN1). By using the extracted features, the performance of different machine learning classifiers, such as k-nearest neighbors (KNN), support vector machine, and decision tree (DT), were compared to find the best algorithm for VIA. After determining the best performing classifying model, it was used to evaluate the screening performance of VIA. RESULTS: An average accuracy of 78%, with a Cohen kappa of 0.571, was observed for the evaluation of the system by four physicians. Through image processing, 240 sliced images were obtained from the cervicogram at each clock position, and five features of VIA were extracted. Among the three models, KNN showed the best performance for finding VIA within holdout 10-fold cross-validation, with an accuracy of 78.3%, area under the curve of 0.807, a specificity of 80.3%, and a sensitivity of 75.0%, respectively. The trained model performed using an unprovided data set resulted in an accuracy of 80.8%, specificity of 84.1%, and sensitivity of 71.9%. Predictions were visualized with intuitive color labels, indicating the normal/abnormal tissue using a circular clock-type segmentation. Calculating the overlapped abnormal tissues between the gold standard and predicted value, the KNN model overperformed the average assessments of physicians for finding VIA. CONCLUSIONS: We explored the potential of the smartphone-based endoscopic VIA as an evaluation technique and used the cervicogram to evaluate normal/abnormal tissue using machine learning techniques. The results of this study demonstrate its potential as a screening tool in low-resource settings.

Smartphone-Based Endoscope System for Advanced Point-of-Care Diagnostics: Feasibility Study.

BACKGROUND: Endoscopic technique is often applied for the diagnosis of diseases affecting internal organs and image-guidance of surgical procedures. Although the endoscope has become an indispensable tool in the clinic, its utility has been limited to medical offices or operating rooms because of the large size of its ancillary devices. In addition, the basic design and imaging capability of the system have remained relatively unchanged for decades. OBJECTIVE: The objective of this study was to develop a smartphone-based endoscope system capable of advanced endoscopic functionalities in a compact size and at an affordable cost and to demonstrate its feasibility of point-of-care through human subject imaging. METHODS: We developed and designed to set up a smartphone-based endoscope system, incorporating a portable light source, relay-lens, custom adapter, and homebuilt Android app. We attached three different types of existing rigid or flexible endoscopic probes to our system and captured the endoscopic images using the homebuilt app. Both smartphone-based endoscope system and commercialized clinical endoscope system were utilized to compare the imaging quality and performance. Connecting the head-mounted display (HMD) wirelessly, the smartphone-based endoscope system could superimpose an endoscopic image to real-world view. RESULTS: A total of 15 volunteers who were accepted into our study were captured using our smartphone-based endoscope system, as well as the commercialized clinical endoscope system. It was found that the imaging performance of our device had acceptable quality compared with that of the conventional endoscope system in the clinical setting. In addition, images captured from the HMD used in the smartphone-based endoscope system improved eye-hand coordination between the manipulating site and the smartphone screen, which in turn reduced spatial disorientation. CONCLUSIONS: The performance of our endoscope system was evaluated against a commercial system in routine otolaryngology examinations. We also demonstrated and evaluated the feasibility of conducting endoscopic procedures through a custom HMD.

Perturbation and differential Monte Carlo methods for measurement of optical properties in a layered epithelial tissue model.

The use of perturbation and differential Monte Carlo (pMC/dMC) methods in conjunction with nonlinear optimization algorithms were proposed recently as a means to solve inverse photon migration problems in regionwise heterogeneous turbid media. We demonstrate the application of pMC/dMC methods for the recovery of optical properties in a two-layer extended epithelial tissue model from experimental measurements of spatially resolved diffuse reflectance. The results demonstrate that pMC/dMC methods provide a rapid and accurate approach to solve two-region inverse photon migration problems in the transport regime, that is, on spatial scales smaller than a transport mean free path and in media where optical scattering need not dominate absorption. The pMC/dMC approach is found to be effective over a broad range of absorption (50 to 400%) and scattering (70 to 130%) perturbations. The recovery of optical properties from spatially resolved diffuse reflectance measurements is examined for different sets of source-detector separation. These results provide some guidance for the design of compact fiber-based probes to determine and isolate optical properties from both epithelial and stromal layers of superficial tissues.

Use of the delta-P1 approximation for recovery of optical absorption, scattering, and asymmetry coefficients in turbid media.

We introduce a robust method to recover optical absorption, reduced scattering, and single-scattering asymmetry coefficients (microa, micro's, g1) of infinite turbid media over a range of (micro's/microa) spanning 3 orders of magnitude. This is accomplished through the spatially resolved measurement of irradiance at source-detector separations spanning 0.25-8 transport mean free paths (l*). These measurements are rapidly processed by a multistaged nonlinear optimization algorithm in which the measured irradiances are compared with predictions given by the delta-P1 variant of the diffusion approximation to the Boltzmann transport equation. The ability of the delta-P1 model to accurately describe radiative transport within media of arbitrary albedo and on spatial scales comparable to l* is the key element enabling the separation of g1 from micro's.

Sampling tissue volumes using frequency-domain photon migration.

We demonstrate the use of Monte Carlo simulations to generate photon scattering density functions (PSDFs) that represent the tissue volume sampled by steady-state and frequency-domain photon migration. We use these results to illustrate how scaling laws can be developed to determine the mean sampling depth of the multiply scattered photons detected by photon migration methods that remain valid outside the bounds of the standard diffusion approximation, i.e., at small source-detector separations and in media where the optical absorption is significant relative to scattering. Using both the PSDF computation and the newly formulated scaling laws, we focus on a comprehensive description of the effects of source modulation frequency, optical absorption, and source-detector separation on the depth of the sampled tissue volume as well as the sensitivity of frequency-domain photon migration measurements to the presence of a localized absorption heterogeneity.