Optical coherence tomography in Barrett's esophagus

Ovarian cancer detection using optical coherence tomography and convolutional neural networks

Ovarian cancer has the sixth-largest fatality rate in the United States among all cancers. A non-surgical assay capable of detecting ovarian cancer with acceptable sensitivity and specificity has yet to be developed. However, such a discovery would profoundly impact the pace of the treatment and improvement to patients' quality of life. Achieving such a solution requires high-quality imaging, image processing, and machine learning to support an acceptably robust automated diagnosis. In this work, we propose an automated framework that learns to identify ovarian cancer in transgenic mice from optical coherence tomography (OCT) recordings. Classification is accomplished using a neural network that perceives spatially ordered sequences of tomograms. We present three neural network-based approaches, namely a VGG-supported feed-forward network, a 3D convolutional neural network, and a convolutional LSTM (Long Short-Term Memory) network. Our experimental results show that our models achieve a favorable performance with no manual tuning or feature crafting, despite the challenging noise inherent in OCT images. Specifically, our best performing model, the convolutional LSTM-based neural network, achieves a mean AUC (± standard error) of 0.81 ± 0.037. To the best of the authors' knowledge, no application of machine learning to analyze depth-resolved OCT images of whole ovaries has been documented in the literature. A significant broader impact of this research is the potential transferability of the proposed diagnostic system from transgenic mice to human organs, which would enable medical intervention from early detection of an extremely deadly affliction.

The Status of Advanced Imaging Techniques for Optical Biopsy of Colonic Polyps

The progressive miniaturization of photonic components presents the opportunity to obtain unprecedented microscopic images of colonic polyps in real time during endoscopy. This information has the potential to act as "optical biopsy" to aid clinical decision-making, including the possibility of adopting new paradigms such as a "resect and discard" approach for low-risk lesions. The technologies discussed in this review include confocal laser endomicroscopy, optical coherence tomography, multiphoton microscopy, Raman spectroscopy, and hyperspectral imaging. These are in different stages of development and clinical readiness, but all show the potential to produce reliable in vivo discrimination of different tissue types. A structured literature search of the imaging techniques for colorectal polyps has been conducted. The significant developments in endoscopic imaging were identified for each modality, and the status of current development was discussed. Of the advanced imaging techniques discussed, confocal laser endomicroscopy is in clinical use and, under optimal conditions with an experienced operator, can provide accurate histological assessment of tissue. The remaining techniques show potential for incorporation into endoscopic equipment and practice, although further component development is needed, followed by robust prospective validation of accuracy. Optical coherence tomography illustrates tissue "texture" well and gives good assessment of mucosal thickness and layers. Multiphoton microscopy produces high-resolution images at a subcellular resolution. Raman spectroscopy and hyperspectral imaging are less developed endoscopically but provide a tissue "fingerprint" which can distinguish between tissue types. Molecular imaging may become a powerful adjunct to other techniques, with its ability to precisely label specific molecules within tissue and thereby enhance imaging.

A Review of New and Emerging Techniques For Optical Diagnosis of Colonic Polyps

BACKGROUND AND AIMS

Endoscopic imaging is a rapidly progressing field and benefits from miniaturization of advanced imaging technologies, which may allow accurate real-time characterization of lesions. The concept of the "optical biopsy" to predict polyp histology has gained prominence in recent years and may become clinically applicable with the advent of new imaging technology. This review aims to discuss current evidence and examine the emerging technologies as applied to the optical diagnosis of colorectal polyps.

METHODS

A structured literature search and review has been carried out of the evidence for diagnostic accuracy of image-enhanced endoscopy and emerging endoscopic imaging technologies. The image-enhanced endoscopy techniques are reviewed, including their basic scientific principles and current evidence for effectiveness. These include the established image-enhancement technologies such as narrow-band imaging, i-scan, and Fuji intelligent chromoendoscopy. More recent technologies including optical enhancement, blue laser imaging, and linked color imaging are discussed. Adjunctive imaging techniques in current clinical use are discussed, such as autofluorescence imaging and endocytoscopy. The emerging advanced imaging techniques are reviewed, including confocal laser endomicroscopy, optical coherence tomography, and Raman spectroscopy.

CONCLUSIONS

Large studies of the established image-enhancement techniques show some role for the optical diagnosis of polyp histology, although results have been mixed, and at present only the technique of narrow-band imaging is appropriate for the diagnosis of low-risk polyps when used by an expert operator. Other image-enhancement techniques will require further study to validate their accuracy but show potential to support the use of a "resect-and-discard" approach to low-risk polyps. New technologies show exciting potential for real-time diagnosis, but further clinical studies in humans have yet to be performed.

Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue

Ovarian cancer has the lowest survival rate among all gynecologic cancers predominantly due to late diagnosis. Early detection of ovarian cancer can increase 5-year survival rates from 40% up to 92%, yet no reliable early detection techniques exist. Optical coherence tomography (OCT) is an emerging technique that provides depth-resolved, high-resolution images of biological tissue in real-time and demonstrates great potential for imaging of ovarian tissue. Mouse models are crucial to quantitatively assess the diagnostic potential of OCT for ovarian cancer imaging; however, due to small organ size, the ovaries must first be separated from the image background using the process of segmentation. Manual segmentation is time-intensive, as OCT yields three-dimensional data. Furthermore, speckle noise complicates OCT images, frustrating many processing techniques. While much work has investigated noise-reduction and automated segmentation for retinal OCT imaging, little has considered the application to the ovaries, which exhibit higher variance and inhomogeneity than the retina. To address these challenges, we evaluate a set of algorithms to segment OCT images of mouse ovaries. We examine five preprocessing techniques and seven segmentation algorithms. While all preprocessing methods improve segmentation, Gaussian filtering is most effective, showing an improvement of 32 % ± 1.2 % . Of the segmentation algorithms, active contours performs best, segmenting with an accuracy of 94.8 % ± 1.2 % compared with manual segmentation. Even so, further optimization could lead to maximizing the performance for segmenting OCT images of the ovaries.

Three-dimensional texture analysis of optical coherence tomography images of ovarian tissue

Ovarian cancer has the lowest survival rate among all gynecologic cancers due to predominantly late diagnosis. Optical coherence tomography (OCT) has been applied successfully to experimentally image the ovaries in vivo; however, a robust method for analysis is still required to provide quantitative diagnostic information. Recently, texture analysis has proved to be a useful tool for tissue characterization; unfortunately, existing work in the scope of OCT ovarian imaging is limited to only analyzing 2D sub-regions of the image data, discarding information encoded in the full image area, as well as in the depth dimension. Here we address these challenges by testing three implementations of texture analysis for the ability to classify tissue type. First, we test the traditional case of extracted 2D regions of interest; then we extend this to include the entire image area by segmenting the organ from the background. Finally, we conduct a full volumetric analysis of the image volume using 3D segmented data. For each case, we compute features based on the Grey-Level Co-occurence Matrix and also by introducing a new approach that evaluates the frequency distribution in the image by computing the energy density. We test these methods on a mouse model of ovarian cancer to differentiate between age, genotype, and treatment. The results show that the 3D application of texture analysis is most effective for differentiating tissue types, yielding an average classification accuracy of 78.6%. This is followed by the analysis in 2D with the segmented image volume, yielding an average accuracy of 71.5%. Both of these improve on the traditional approach of extracting square regions of interest, which yield an average classification accuracy of 67.7%. Thus, applying texture analysis in 3D with a fully segmented image volume is the most robust approach to quantitatively characterizing ovarian tissue.

Future of diagnosing neoplasia in Barrett's esophagus: volumetric laser endomicroscopy

Esophageal adenocarcinoma (EAC) is one of the deadliest carcinoma faced by gastroenterologists. Any insult to esophagus that causes replacement of normal squamous epithelium with columnar intestinal epithelium is labelled as the initiating event of the metaplasia-neoplasia sequence. Barrett's esophagus is the precursor to EAC. Currently, endoscopically obtained biopsies are used to detect neoplastic changes in patients with Barrett's esophagus (BE); however, it is not cost effective and hence a better screening modality is needed. Volumetric laser endomicroscopy (VLE) has been under study for the past few years and has shown promising results to overcome the shortcoming faced in the biopsy samplings. It is a second-generation optical coherence tomography (OCT) that provides high-resolution cross-sectional imaging of the esophageal mucosa using near-infrared light. The principle is similar to endosonography, but image formation in OCT depends on variations in the reflection of light from different tissue layers rather than ultrasonic waves.

Endoscopic submucosal dissection and endoscopic mucosal resection for early stage esophageal cancer

Mortality from esophageal cancer remains high despite advances in medical therapy. Although the incidence of squamous cell carcinoma of the esophagus remains unchanged, the incidence of the esophageal adenocarcinoma has increased over time. Gastroesophageal reflux disease (GERD and obesity are contributing factors to the development of Barrett's esophagus and subsequent development of adenocarcinoma. Early recognition of the disease can lead to resection of esophageal cancer prior to the development of lymphovascular invasion. Various modalities have been implemented to aid identification of precancerous lesions and early esophageal cancer. Chromoendoscopy, narrowband imaging and endoscopic ultrasound examination are typically used for evaluating early esophageal lesions. Recently, confocal laser endomicroscopy (CLE) and volumetric laser scanning were implemented with promising results. Endoscopic management of early esophageal cancer may be done using endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD). Both techniques allow resection of the mucosa (and possibly a portion of the submucosa) containing the early tumor without interruption of deeper layers. A submucosal injection creating a cushion coupled with snare resection or cap assisted mucosal suction followed by ligation and snare resection are the most common techniques of EMR. EMR can remove lesions less than 2 cm in size en bloc. Larger lesions may require resection in piecemeal fashion. This may limit assessment of the margins of the lesion and orienting the lesion's border. ESD offers en bloc dissection of the lesion regardless of its size. ESD is performed with specialized needle knives, which allow incision followed by careful dissection of the lesion within the submucosal layer. Tumor recurrence after ESD is rare but the technique is labor intensive and has an increased risk of perforation. Esophageal stenosis remains a concern after extensive EMR or ESD. Dilation with balloon or stent placement is usually sufficient to treat post-resection stenosis.

Prostate cancer diagnosis by optical coherence tomography: First results from a needle based optical platform for tissue sampling

The diagnostic accuracy of Optical Coherence Tomography (OCT) based optical attenuation coefficient analysis is assessed for the detection of prostate cancer. Needle-based OCT-measurements were performed on the prostate specimens. Attenuation coefficients were determined by an earlier described in-house developed software package. The mean attenuation coefficients (benign OCT data; malignant OCT data; p-value Mann-Whitney U test) were: (3.56 mm(-1) ; 3.85 mm(-1) ; p < 0.0001) for all patients combined. The area under the ROC curve was 0.64. In order to circumvent the effect of histopathology mismatching, we performed a sub-analysis on only OCT data in which tumor was visible in two subsequent histopathological prostate slices. This analysis could be performed in 3 patients. The mean attenuation coefficients (benign OCT data; malignant OCT data; p-value Mann-Whitney U test) were: (3.23 mm(-1) ; 4.11 mm(-1) ; p < 0.0001) for all patients grouped together. The area under the ROC curve was 0.89. Functional OCT of the prostate has shown to differentiate between cancer and healthy prostate tissue. The optical attenuation coefficient in malignant tissue was significantly higher in malignant tissue compared to benign prostate tissue. Further studies are required to validate these initial results in a larger group of patients with a more tailored histopathology matching protocol.