Volumetric Laser Endomicroscopy Scans Research Articles

Artificial intelligence-enhanced volumetric laser endomicroscopy improves dysplasia detection in Barrett’s esophagus in a randomized cross-over study

Volumetric laser endomicroscopy (VLE) is an advanced endoscopic imaging tool that can improve dysplasia detection in Barrett’s esophagus (BE). However, VLE scans generate 1200 cross-sectional images that can make interpretation difficult. The impact of a new VLE artificial intelligence algorithm called Intelligent Real-time Image Segmentation (IRIS) is not well-characterized. This is a randomized prospective cross-over study of BE patients undergoing endoscopy who were randomized to IRIS-enhanced or unenhanced VLE first followed by the other (IRIS-VLE vs. VLE-IRIS, respectively) at expert BE centers. The primary outcome was image interpretation time, which served as a surrogate measure for ease of interpretation. The secondary outcome was diagnostic yield of dysplasia for each imaging modality. 133 patients were enrolled. 67 patients were randomized to VLE-IRIS and 66 to IRIS-VLE. Total interpretation time did not differ significantly between groups (7.8 min VLE-IRIS vs. 7 min IRIS-VLE, P = 0.1), however unenhanced VLE interpretation time was significantly shorter in the IRIS-VLE group (2.4 min vs. 3.8 min, P < 0.01). When IRIS was used first, 100% of dysplastic areas were identified, compared with 76.9% when VLE was the first interpretation modality (P = 0.06). IRIS-enhanced VLE reduced the time of subsequent unenhanced VLE interpretation, suggesting heightened efficiency and improved dysplasia detection. It was also able to identify all endoscopically non-visible dysplastic areas.

Prospective development and validation of a volumetric laser endomicroscopy computer algorithm for detection of Barrett’s neoplasia

Volumetric laser endomicroscopy (VLE) is an advanced imaging modality used to detect Barrett's esophagus (BE) dysplasia. However, real-time interpretation of VLE scans is complex and time-consuming. Computer-aided detection (CAD) may help in the process of VLE image interpretation. Our aim was to train and validate a CAD algorithm for VLE-based detection of BE neoplasia. The multicenter, VLE PREDICT study, prospectively enrolled 47 patients with BE. In total, 229 nondysplastic BE and 89 neoplastic (high-grade dysplasia/esophageal adenocarcinoma) targets were laser marked under VLE guidance and subsequently underwent a biopsy for histologic diagnosis. Deep convolutional neural networks were used to construct a CAD algorithm for differentiation between nondysplastic and neoplastic BE tissue. The CAD algorithm was trained on a set consisting of the first 22 patients (134 nondysplastic BE and 38 neoplastic targets) and validated on a separate test set from patients 23 to 47 (95 nondysplastic BE and 51 neoplastic targets). The performance of the algorithm was benchmarked against the performance of 10 VLE experts. Using the training set to construct the algorithm resulted in an accuracy of 92%, sensitivity of 95%, and specificity of 92%. When performance was assessed on the test set, accuracy, sensitivity, and specificity were 85%, 91%, and 82%, respectively. The algorithm outperformed all 10 VLE experts, who demonstrated an overall accuracy of 77%, sensitivity of 70%, and specificity of 81%. We developed, validated, and benchmarked a VLE CAD algorithm for detection of BE neoplasia using prospectively collected and biopsy-correlated VLE targets. The algorithm detected neoplasia with high accuracy and outperformed 10 VLE experts. (The Netherlands National Trials Registry (NTR) number: NTR 6728.).

Multicenter study on the diagnostic performance of multiframe volumetric laser endomicroscopy targets for Barrett's esophagus neoplasia with histopathology correlation.

Volumetric laser endomicroscopy (VLE) has been shown to improve detection of early neoplasia in Barrett’s esophagus (BE). However, diagnostic performance using histopathology-correlated VLE regions of interest (ROIs) has not been adequately studied. We evaluated the diagnostic accuracy of VLE assessors for identification of early BE neoplasia in histopathology-correlated VLE ROIs. In total, 191 ROIs (120 nondysplastic and 71 neoplastic) from 50 BE patients were evaluated in a random order using a web-based module. All ROIs contained histopathology correlations enabled by VLE laser marking. Assessors were blinded to endoscopic BE images and histology. ROIs were first scored as nondysplastic or neoplastic. Level of confidence was assigned to the predicted diagnosis. Outcome measures were: (i) diagnostic performance of VLE assessors for identification of BE neoplasia in all VLE ROIs, defined as accuracy, sensitivity, and specificity; (ii) diagnostic performance of VLE assessors for only high level of confidence predictions; and (iii) interobserver agreement. Accuracy, sensitivity, and specificity for BE neoplasia identification were 79% (confidence interval [CI], 75–83), 75% (CI, 71–79), and 81% (CI, 76–86), respectively. When neoplasia was identified with a high level of confidence, accuracy, sensitivity, and specificity were 88%, 83%, and 90%, respectively. The overall strength of interobserver agreement was fair (k = 0.29). VLE assessors can identify BE neoplasia with reasonable diagnostic accuracy in histopathology-correlated VLE ROIs, and accuracy is enhanced when BE neoplasia is identified with high level of confidence. Future work should focus on renewed VLE image reviewing criteria and real-time automatic assessment of VLE scans.

Mo1256 EPITHELIAL GLAND BURDEN DETERMINED BY VOLUMETRIC LASER ENDOMICROSCOPY IS ASSOCIATED WITH COMPLETE REMISSION OF INTESTINAL METAPLASIA FOLLOWING ENDOSCOPIC ERADICATION THERAPY FOR BARRETT’S ESOPHAGUS

Endoscopic eradication of Barrett’s esophagus (BE) aims to achieve complete remission of intestinal metaplasia (CRIM), defined by absence of visible columnar mucosa on high-definition white light (HDWL) endoscopy and negative biopsies for intestinal metaplasia (IM). Given sampling error associated with random biopsies, some experts recommend two consecutive negative endoscopies to establish CRIM. Volumetric laser endomicroscopy (VLE) is an advanced imaging modality capable of comprehensive high-resolution imaging of BE. Epithelial glands are a VLE feature associated with BE. We hypothesize that VLE epithelial gland burden in the gastroesophageal junction (GEJ) is associated with the presence of IM in patients with BE following endoscopic eradication therapy. Patients with BE with absence of visible columnar mucosa on HDWL endoscopy following endoscopic eradication therapy underwent VLE imaging. Random surveillance biopsies were obtained as per standard protocol. VLE scans were analyzed with intelligent real-time image segmentation (IRIS), an artificial intelligence software that superimposes features of interest (signal intensity, mucosal layering, epithelial glands) onto an en-face map of the esophagus. The GEJ on IRIS en-face scans was defined as the transition between squamous epithelium (presence of layering, normal signal intensity) and cardia (lack of layering, increased signal intensity) and demarcated using image processing software ImageJ. We previously showed that the surface area occupied by epithelial glands correlates with manual gland quantification. The GEJ demarcation was transferred to an en-face map of epithelial glands and the surface area (gland burden) was measured 1 cm proximal and 1 cm distal to the GEJ for each VLE scan (Figure 1) The mean gland surface area of VLE scans from patients with IM negative biopsies was compared to those with IM positive biopsies using the Student t-test. This study was funded by the ASGE Endoscopy Research Award. A total of 123 BE patients without visible BE on HDWL endoscopy following endoscopic therapy underwent VLE. Random biopsies were negative for IM in 87 and positive in 36 patients. Highest grade of histology was non-dysplastic BE (N=32), indefinite (N=2), low-grade (N=2). The mean (SD) gland surface area in the GEJ was significantly higher in patients with IM positive biopsies compared to patients with IM negative biopsies [10.1(9.9)mm2 vs. 6.5(5.5)mm2, p=0.02] (Figure 2) Systematic quantification of epithelial gland burden in the GEJ using VLE can predict presence of IM in patients with BE following endoscopic eradication therapy. VLE may have a role in confirming true CRIM and reducing number of required surveillance endoscopies.Mean gland surface area measured in patients with absence of visible columnar mucosa on high-definition white light endoscopy and negative surveillance biopsies for IM (group 0) and positive surveillance biopsies for IM (group 1)View Large Image Figure ViewerDownload Hi-res image Download (PPT)

Barrett’s Epithelial Thickness, Assessed by Volumetric Laser Endomicroscopy, Is Associated With Response to Radiofrequency Ablation

Radiofrequency ablation (RFA) is the most common treatment for flat Barrett's esophagus (BE), but reasons for varying outcomes are poorly understood. A recently developed contrast-enhancement algorithm allows reliable measurement of Barrett's epithelial thickness (BET) from volumetric laser endomicroscopy (VLE) images and correlation with response to RFA. Using this algorithm, we investigated whether patients with thicker Barrett's mucosa are less likely to respond to RFA. In the future, this algorithm may guide choice of RFA dosing or endoscopic resection. We performed a retrospective analysis on all patients with BE who received a baseline VLE scan between May 2015 and October 2016, followed by RFA and 1 follow-up exam, from 14 institutions participating in the United States VLE registry. We measured BET on equidistant locations by estimating the distance between the esophageal surface and the superficial edge of the deepest lamina propria. The primary outcome variable was the percentage reduction in Prague length; secondary outcome variables were complete remission of intestinal metaplasia (CRIM) and presence of strictures after 12 months. Images from 61 patients were included in our final analysis. Mean BET per patient ranged from 224 μm to 705 μm. A 100 μm thicker mean BET per patient resulted in a 12% lower response to treatment, measured by a reduction of Prague length (P = .03), after adjustment for confounders. We found an association between mean BET and CRIM, but not with stricture formation. Based on measurements on contrast-enhanced VLE images, we found that BET correlates with response to RFA. For clinical implementation, larger studies with a standardized follow-up and development of computer-aided image analysis systems are needed.

Deep principal dimension encoding for the classification of early neoplasia in Barrett's Esophagus with volumetric laser endomicroscopy

Barrett cancer is a treatable disease when detected at an early stage. However, current screening protocols are often not effective at finding the disease early. Volumetric Laser Endomicroscopy (VLE) is a promising new imaging tool for finding dysplasia in Barrett's esophagus (BE) at an early stage, by acquiring cross-sectional images of the microscopic structure of BE up to 3-mm deep. However, interpretation of VLE scans is difficult for medical doctors due to both the size and subtlety of the gray-scale data. Therefore, algorithms that can accurately find cancerous regions are very valuable for the interpretation of VLE data. In this study, we propose a fully-automatic multi-step Computer-Aided Detection (CAD) algorithm that optimally leverages the effectiveness of deep learning strategies by encoding the principal dimension in VLE data. Additionally, we show that combining the encoded dimensions with conventional machine learning techniques further improves results while maintaining interpretability. Furthermore, we train and validate our algorithm on a new histopathologically validated set of in-vivo VLE snapshots. Additionally, an independent test set is used to assess the performance of the model. Finally, we compare the performance of our algorithm against previous state-of-the-art systems. With the encoded principal dimension, we obtain an Area Under the Curve (AUC) and F1 score of 0.93 and 87.4% on the test set respectively. We show this is a significant improvement compared to the state-of-the-art of 0.89 and 83.1%, respectively, thereby demonstrating the effectiveness of our approach.

Ensemble of Deep Convolutional Neural Networks for Classification of Early Barrett’s Neoplasia Using Volumetric Laser Endomicroscopy

Barrett’s esopaghagus (BE) is a known precursor of esophageal adenocarcinoma (EAC). Patients with BE undergo regular surveillance to early detect stages of EAC. Volumetric laser endomicroscopy (VLE) is a novel technology incorporating a second-generation form of optical coherence tomography and is capable of imaging the inner tissue layers of the esophagus over a 6 cm length scan. However, interpretation of full VLE scans is still a challenge for human observers. In this work, we train an ensemble of deep convolutional neural networks to detect neoplasia in 45 BE patients, using a dataset of images acquired with VLE in a multi-center study. We achieve an area under the receiver operating characteristic curve (AUC) of 0.96 on the unseen test dataset and we compare our results with previous work done with VLE analysis, where only AUC of 0.90 was achieved via cross-validation on 18 BE patients. Our method for detecting neoplasia in BE patients facilitates future advances on patient treatment and provides clinicians with new assisting solutions to process and better understand VLE data.

Interpretation of volumetric laser endomicroscopy in Barrett's esophagus using image enhancement software.

Image interpretation of Barrett's esophagus (BE) with volumetric laser endomicroscopy (VLE) can be enhanced by image processing software that highlights established features using a color-graded scale (intelligent real-time image segmentation, IRIS). This study aims to provide a description of IRIS features of various gastroesophageal tissue types using histologic correlation. A database of 80 VLE laser-marked targets with histologic correlation was reviewed for various tissue types. IRIS was applied off-line to the VLE scans, laser-marked targets were identified, and feature review was performed. Squamous epithelium targets (N = 7) showed IRIS layered architecture with lack of surface hyper-reflectivity and epithelial glands. Gastric cardia targets (N = 10) showed absent layering (100%) and surface hyper-reflectivity with epithelial glands (40%). Nondysplastic BE targets (N = 39) showed surface hyper-reflectivity (64%), epithelial glands (51%), and lack of layering (74%). Targets of BE with early neoplasia (N = 24), showed surface hyper-reflectivity (96%), epithelial glands (67%), and lack of layering (96%). IRIS features that characterize each tissue type appear to mirror the nonenhanced VLE counterparts that define them.

Staging of T1 esophageal adenocarcinoma with volumetric laser endomicroscopy: a feasibility study.

Background and study aims Precise staging in T1 esophageal adenocarcinoma (EAC) is critical in determining candidacy for curative endoscopic resection. High-frequency endoscopic ultrasound (EUS) has demonstrated suboptimal accuracy in T1 EAC staging due to insufficient spatial resolution. Volumetric laser endomicroscopy (VLE) allows for high-resolution wide-field visualization of the esophageal microstructure. We aimed to investigate the role of VLE in staging T1 EAC. Patients and methods Patients undergoing endoscopic mucosal resection (EMR) were prospectively enrolled and only T1 EAC cases were included. EMR specimens were imaged using second-generation VLE immediately after resection. VLE images were analyzed for signal intensity by depth and signal attenuation (dB/mm) in both cross-sectional and en-face orientation. A decision tree model was constructed to combine measured VLE parameters and delineate diagnostic thresholds. Results Thirty EMR scans were obtained – 15 T1a specimens from 9 patients and 15 T1b specimens from 11 patients. T1b specimen VLE scans exhibited higher signal intensity ( P < 0.0001) and higher signal attenuation compared to T1a specimens ( P = 0.03). A combination of signal attenuation and signal intensity at 150 µm depth yielded optimal diagnostic thresholds and an area under the curve (AUC) of 0.77. VLE signal attenuation was significantly associated with grade of differentiation, irrespective of EAC stage. Conclusions VLE signal intensity and signal attenuation are quantitatively distinct in T1a and T1b EAC and associated with grade of differentiation. This is the first study examining the role of VLE for staging of T1 EAC and demonstrates promising diagnostic performance. With further in vivo validation, VLE may serve a role in staging superficial EAC.

Measuring Barrett\u2019s Epithelial Thickness with Volumetric Laser Endomicroscopy as a Biomarker to Guide Treatment

BackgroundRadiofrequency ablation (RFA) treatment outcomes vary for unknown reasons. One hypothesis is that variations in Barrett’s epithelial thickness (BET) are associated with reduced RFA efficacy for thicker BET and strictures for thinner BET. Volumetric laser endomicroscopy (VLE) is an imaging modality that acquires high-resolution, depth-resolved images of BE. However, the attenuation of light by tissue and the lack of layering in Barrett’s tissue challenge BET measurements and the study of relationships between thickness and RFA outcomes. We aimed to quantify BET and compared the reliability of standard and contrast-enhanced VLE images.MethodsBaseline VLE scans from BE patients without prior ablative therapy and a Prague (M) length of > 1 cm were obtained from the US VLE Registry. An algorithm was applied to the VLE images to flatten the mucosal surface and enhance the contrast of different esophageal wall layers. Subsequently, BET was measured by two independent VLE readers using both contrast- and non-contrast-enhanced datasets. In order to validate these adjusted images, intra- and interobserver agreements were calculated.ResultsVLE scans from fifty-seven patients were included in this study. BET was measured at eight equidistant locations on the selected cross-sectional images at 0.5 cm intervals from the GEJ to the proximal-most extent of BE. The intra-observer coefficients of the two readers for the contrast-enhanced images were 0.818 (95% CI 0.798–0.836) and 0.890 (95% CI 0.878–0.900). The interobserver agreement for the contrast-enhanced images (0.880; 95% CI 0.867–0.891) was significantly better than for the original images (0.778; 95% CI 0.754–0.799).ConclusionWe developed an algorithm that improves VLE visualization of the mucosal layers of the esophageal wall and enables rapid and reliable measurement of BET. Interobserver variability measurements were significantly reduced when using contrast enhancement. Studies are underway to correlate BET with treatment response.

Use of volumetric laser endomicroscopy for determining candidates for endoscopic therapy in superficial esophageal squamous cell carcinoma.

Accurate staging of superficial esophageal squamous cell cancer (ESCC) for endoscopic therapy is challenging. Optical coherence tomography (OCT) has been shown to be superior to high-resolution endoscopic ultrasound (EUS). Volumetric Laser Endomicroscopy (VLE), a second-generation OCT, has recently become commercially available. To assess if VLE can determine which patients with superficial ESSC can undergo endoscopic therapy. This is a multi-center retrospective study. Patients were included if (a) they had visible ESCC, (b) they underwent VLE and EUS for staging, and c) if superficial disease was suspected then endoscopic resection had to be performed to have accurate histology to compare the VLE scan to. VLE scans were then compared to the gold standard: histology for superficial disease and EUS for disease T1b and greater. Seventeen patients were included with the following disease: squamous intraepithelial neoplasia (4 patients), T1a disease (6 patients), T1b (2 patients), T2 disease (2 patients) and T3 disease (3 patients). VLE was able to distinguish superficial disease, defined as disease limited up to the lamina propria, from non-superficial disease in all cases. VLE may be able to determine which ESCC patients are candidates for endoscopic therapy. Prospective studies are needed to confirm this.

Volumetric laser endomicroscopy interpretation and feature analysis in dysplastic Barrett's esophagus.

Volumetric laser endomicroscopy (VLE) is used to identify Barrett's esophagus (BE) dysplasia. Selection of a dysplastic region of interest (ROI) can be challenging due to feature variability across a large amount of data. The degree of agreement among VLE users in selecting a ROI has not been studied. High-definition videos that divided a VLE scan from 18 patients with biopsy-proven BE dysplasia into 1-cm segments were reviewed using a four-quadrant grid superimposed for systematic interpretation. VLE scans were selected based on image quality and appropriate visualization of BE epithelium. Four experienced VLE users rated each quadrant as dysplastic or non-dysplastic. For quadrants rated as dysplastic, reviewers selected a single timeframe with representative features. A high-degree of agreement among reviewers was defined as ≥75% agreement on the quadrant diagnosis and ≥50% agreement on selected timeframe (±2s). Thirty-one videos, each 32s in length, comprising 124 quadrants were reviewed. There was high-agreement among reviewers in 99 (80%) quadrants, of which 68 (69%) were rated as dysplastic. Compared with quadrants rated as non-dysplastic, ROIs of quadrants rated as dysplastic contained a higher number of epithelial glands (12.7 vs 1.2, P<0.001) with atypical architecture (54 vs 1, P<0.001). A statistically significant difference was observed between the signal intensity profiles of quadrants rated as dysplastic and quadrants rated as non-dysplastic (P=0.004). This study highlights that experienced VLE users can identify ROIs with high-degree of agreement. Selected ROIs contained VLE features associated with BE dysplasia.

Tu1120 - Quantitative Image Analysis of Volumetric Laser Endomicroscopy Scans Predicts Tumor Stage and is Associated with Grade of Differentiation in T1 Esophageal Adenocarcinoma

Tu1120 - Quantitative Image Analysis of Volumetric Laser Endomicroscopy Scans Predicts Tumor Stage and is Associated with Grade of Differentiation in T1 Esophageal Adenocarcinoma

Predictive features for early cancer detection in Barrett's esophagus using Volumetric Laser Endomicroscopy

The incidence of Barrett cancer is increasing rapidly and current screening protocols often miss the disease at an early, treatable stage. Volumetric Laser Endomicroscopy (VLE) is a promising new tool for finding this type of cancer early, capturing a full circumferential scan of Barrett's Esophagus (BE), up to 3-mm depth. However, the interpretation of these VLE scans can be complicated, due to the large amount of cross-sectional images and the subtle grayscale variations. Therefore, algorithms for automated analysis of VLE data can offer a valuable contribution to its overall interpretation. In this study, we broadly investigate the potential of Computer-Aided Detection (CADe) for the identification of early Barrett's cancer using VLE. We employ a histopathologically validated set of ex-vivo VLE images for evaluating and comparing a considerable set of widely-used image features and machine learning algorithms. In addition, we show that incorporating clinical knowledge in feature design, leads to a superior classification performance and additional benefits, such as low complexity and fast computation time. Furthermore, we identify an optimal tissue depth for classification of 0.5–1.0 mm, and propose an extension to the evaluated features that exploits this phenomenon, improving their predictive properties for cancer detection in VLE data. Finally, we compare the performance of the CADe methods with the classification accuracy of two VLE experts. With a maximum Area Under the Curve (AUC) in the range of 0.90–0.93 for the evaluated features and machine learning methods versus an AUC of 0.81 for the medical experts, our experiments show that computer-aided methods can achieve a considerably better performance than trained human observers in the analysis of VLE data.

Suspicious-appearing Epithelial Glands on Volumetric Laser Endomicroscopy Predict the Presence of Dysplasia During Surveillance of Barrettʼs Esophagus

Introduction: Epithelial dysplasia (ED) within Barrett's esophagus (BE) increases the risk of progression to adenocarcinoma. Traditional surveillance using 4-quadrant forceps biopsies every 1-2 cm leaves most tissue unsampled, decreasing diagnostic accuracy. Alternate sampling methods and advanced imaging techniques, which evaluate a greater percentage of tissue, may address this deficiency. Volumetric Laser Endomicroscopy (VLE) provides real-time cross-sectional esophageal imaging to a resolution of 7 microns and a depth of 3 mm. Suspicious-appearing glands (SG) within BE epithelium may predict ED. Our aim was to evaluate whether the number of SG within a BE segment, and/or within an area of concern (AOC), predicts the presence of ED. Methods: All endoscopies performed between 3/2015 and 1/2017 for inspection of non-treated BE by a single endoscopist with expertise in VLE interpretation were included. Demographics and BE-related results, including Prague scores, number of SG within each AOC, and SG characteristics (shape, presence of necrosis) were recorded, along with the highest dysplasia grade from all tissue sampled. Findings were deidentified and aggregated for analysis. Results: A total of 127 cases met inclusion criteria. Mean Prague scores were C=1.48 cm (0-17 cm), M=2.83 cm (0-18 cm). Of these, 37 patients (29%) had dysplasia confirmed on pathology. In non-dysplastic cases, SG were absent in 60% (53/90), whereas SG were found in 78% (29/37) of dysplastic cases. The relative risk for dysplasia with SG present was 3.35 (95% CI 1.66-6.75, P=0.0007), while the negative predictive value for dysplasia with SG absent was 86.9% (95% CI 77.8%-92.6%). The likelihood of finding dysplasia increased based on the total number of SG present in a BE segment (12.5% with 1 SG, 64.7% if 10+ SG), as did the number of SG within a single AOC (39.2% with 1-2 SG, 74.1% with 6+ SG). Specific characteristics of each SG did not affect likelihood of dysplasia. Conclusion: These results suggest the presence of SG on a VLE scan is strongly associated with dysplastic findings on biopsy. Increased numbers of SG, both overall and within a specific AOC, appear to increase the likelihood of confirming dysplasia as well. Combined with other VLE features such as layering effacement and hyperscattering changes, the presence of SG may improve risk stratification of BE surveillance patients. Further research on biopsy-correlated datasets is necessary to identify the diagnostic accuracy of these VLE features.Table: Table. Likelihood of Finding Dysplasia on Biopsy Based on Total Number of Suspicious-Appearing Epithelial GlandsTable: Table. Likelihood of Finding Dysplasia on Biopsy Based on Maximum Number of Suspicious-Appearing Epithelial Glands Within An Area of Concern

Barrettʼs Esophagus Lesion Identification With Volumetric Laser Endomicroscopy: Interobserver Agreement Between Expert and Novice Users

Introduction: Volumetric Laser Endomicroscopy (VLE) is used to detect Barrett's esophagus (BE) dysplasia. VLE features of glandular atypia, increased surface signal intensity, and lack of layering are associated with BE dysplasia. Experienced VLE users show high interobserver agreement in identifying a region of interest (ROI) containing these features. The accuracy of novice users in dysplastic ROI identification is unknown. The study aim was to determine the interobserver agreement between novice and experienced VLE users for dysplastic ROI identification. Methods: 18 VLE scans (12 high-grade dysplasia; 6 intramucosal adenocarcinoma) from a US based multicenter registry were selected. To optimize interpretation, scans were divided into videos corresponding to 1 cm of circumferential BE, representing ˜200 cross-sectional images. Each video was divided into four quadrants. For each quadrant, reviewers were asked to select the timeframe with the most representative features of dysplasia. If a quadrant contained no features of dysplasia, reviewers were asked to rate it as non-dysplastic BE. Expert VLE raters had experience reviewing >50 VLE scans. Novice VLE raters were endoscopists with no VLE experience who were presented a 20 minute tutorial of images/videos on VLE features of BE dysplasia. High agreement among VLE reviewers in each group (experts vs. novice) was defined by ≥75% agreement on the quadrant diagnosis (dysplastic vs. non-dysplastic) and ≥50% agreement on timeframe (±5 sec). Results: A total of 31 videos (124 quadrants) were reviewed by 4 expert and 4 novice VLE users. There was high agreement among expert users in 95 (77%) quadrants of which 61 (64%) quadrants were rated as dysplastic. There was high agreement among novice users in 88 (71%) quadrants of which 60 (68%) quadrants were rated as dysplastic. Expert and novice users agreed on the diagnosis of dysplasia in 44 (56%) of 78 quadrants. 17 (22%) quadrants were rated by experts as dysplastic and by novice users as non-dysplastic. 17 (22%) quadrants were rated by novice users as dysplastic and experts as non-dysplastic. There was no interobserver agreement between novice and expert groups in identifying a ROI concerning for dysplasia (kappa -0.27). Conclusion: There was no agreement in identification of dysplastic ROIs with VLE between novice and expert users, highlighting the importance of additional training of novice users in the use of this technology.

Quantitative attenuation analysis for identification of early Barrett's neoplasia in volumetric laser endomicroscopy.

Early neoplasia in Barrett’s esophagus (BE) is difficult to detect. Volumetric laser endomicroscopy (VLE) incorporates optical coherence tomography, providing a circumferential scan of the esophageal wall layers. The attenuation coefficient (μVLE) quantifies decay of detected backscattered light versus depth, and could potentially improve BE neoplasia detection. The aim is to investigate feasibility of μVLE for identification of early BE neoplasia. In vivo and ex vivo VLE scans with histological correlation from BE patients ± neoplasia were used. Quantification by μVLE was performed manually on areas of interest (AoIs) to differentiate neoplasia from nondysplastic (ND)BE. From ex vivo VLE scans from 16 patients (13 with neoplasia), 68 AoIs were analyzed. Median μVLE values (mm−1) were 3.7 [2.1 to 4.4 interquartile range (IQR)] for NDBE and 4.0 (2.5 to 4.9 IQR) for neoplasia, not statistically different (p=0.82). Fourteen in vivo scans were used: nine from neoplastic and five from NDBE patients. Median μVLE values were 1.8 (1.5 to 2.6 IQR) for NDBE and 2.1 (1.9 to 2.6 IQR) for neoplasia, with no statistically significant difference (p=0.37). In conclusion, there was no significant difference in μVLE values in VLE scans from early neoplasia versus NDBE. Future studies with a larger sample size should explore other quantitative methods for detection of neoplasia during BE surveillance.

Computer-aided detection of early Barrett’s neoplasia using volumetric laser endomicroscopy

Volumetric laser endomicroscopy (VLE) is an advanced imaging system that provides a near-microscopic resolution scan of the esophageal wall layers up to 3-mm deep. VLE has the potential to improve detection of early neoplasia in Barrett's esophagus (BE). However, interpretation of VLE images is complex because of the large amount of data that need to be interpreted in real time. The aim of this study was to investigate the feasibility of a computer algorithm to identify early BE neoplasia on exvivo VLE images. We used 60 VLE images from a database of high-quality exvivo VLE-histology correlations, obtained from BE patients ± neoplasia (30 nondysplastic BE [NDBE] and 30 high-grade dysplasia/early adenocarcinoma images). VLE features from a recently developed clinical VLE prediction score for BE neoplasia served as input for the algorithm: (1) higher VLE surface than subsurface signal and (2) lack of layering. With this input, novel clinically inspired algorithm features were developed, based on signal intensity statistics and grayscale correlations. For comparison, generic image analysis methods were examined for their performance to detect neoplasia. For classification of the images in the NDBE or neoplastic group, several machine learning methods were evaluated. Leave-1-out cross-validation was used for algorithm validation. Three novel clinically inspired algorithm features were developed. The feature "layering and signal decay statistics" showed the optimal performance compared with the other clinically features ("layering" and "signal intensity distribution") and generic image analyses methods, with an area under the receiver operating characteristic curve (AUC) of .95. Corresponding sensitivity and specificity were 90% and 93%, respectively. In addition, the algorithm showed a better performance than the clinical VLE prediction score (AUC .81). This is the first study in which a computer algorithm for BE neoplasia was developed based on VLE images with direct histologic correlates. The algorithm showed good performance to detect BE neoplasia in exvivo VLE images compared with the performance of a recently developed clinical VLE prediction score. This study suggests that an automatic detection algorithm has the potential to assist endoscopists in detecting early neoplasia on VLE. Future studies on invivo VLE scans are needed to further validate the algorithm.

Use of volumetric laser endomicroscopy for dysplasia detection at the gastroesophageal junction and gastric cardia.

AIMTo determine specific volumetric laser endomicroscopy (VLE) imaging features associated with neoplasia at the gastroesophageal junction (GEJ) and gastric cardia.METHODSDuring esophagogastroduodenoscopy for patients with known or suspected Barrett’s esophagus, VLE was performed before biopsies were taken at endoscopists’ discretion. The gastric cardia was examined on VLE scan from the GEJ (marked by top of gastric folds) to 1 cm distal from the GEJ. The NinePoints VLE console was used to analyze scan segments for characteristics previously found to correlate with normal or abnormal mucosa. Glands were counted individually. Imaging features identified on VLE scan were correlated with biopsy results from the GEJ and cardia region.RESULTSThis study included 34 cases. Features characteristic of the gastric cardia (gastric rugae, gastric pit architecture, poor penetration) were observed in all (100%) scans. Loss of classic gastric pit architecture was common and there was no difference between those with neoplasia and without (100% vs 74%, P = NS). The abnormal VLE feature of irregular surface was more often seen in patients with neoplasia than those without (100% vs 18%, P < 0.0001), as was heterogeneous scattering (86% vs 41%, P < 0.005) and presence of anomalous glands (100% vs 59%, P < 0.05). The number of anomalous glands did not differ between individual histologic subgroups (ANOVA, P = NS).CONCLUSIONThe transition from esophagus to gastric cardia is reliably identified on VLE. Histologically abnormal cardia mucosa produces abnormal VLE features. Optical coherence tomography algorithms can be expanded for use at the GEJ/cardia.

Hidden in Plain Sight: The Value of Ex Vivo Volumetric Laser Endomicroscopy in Confirming Neoplastic Changes Within a Barrettʼs Nodule

Volumetric Laser Endomicroscopy (VLE) utilizes next generation optical coherence tomography to produce high-resolution cross-sectional imaging via a swept source laser. Usage to date has focused on the evaluation of Barrett's esophagus (BE), specifically identification of features consistent with dysplasia and neoplasia. A 66 year old male was referred for evaluation and treatment of BE found by his local gastroenterologist to have developed multi-focal low grade dysplasia between surveillance endoscopies in 2013 and 2016. The patient was scheduled for upper endoscopy with mapping BE biopsies. Key findings included a 3 cm hiatal hernia and a BE segment classified as C2-M3 by Prague criteria. Within this segment was a subcm nodule located 1 cm proximal to the top of the gastric folds (Figure 1). VLE of the distal esophagus revealed diffuse loss of normal wall layer architecture, consistent with BE, along with a focal area of abnormal epithelial glands suspicious for dysplasia located on the opposite wall from the nodule. The region near the nodule contained a topographical surface, but no clear suspicious glands were appreciated. The nodule then was resected using a cap-band-ligator with snare mucosectomy technique. An ex vivo VLE scan was performed on the nodule. Images demonstrated multiple areas of markedly abnormal-appearing epithelial glands with some containing what appeared to be necrotic material. These changes were suspicious for high grade dysplasia (Figure 2).Figure 1Figure 2Figure 3Pathology from the resected specimen showed extensive BE with high grade dysplasia. Comparison of the tissue sample and the ex vivo VLE images confirmed a very similar morphologic appearance in both specimens. This case highlights how the limitations of in vivo VLE when nodularity is present to distort normal esophageal anatomy can be overcome through use of the technology following endoscopic resection. The topographical nature of the nodule prevented optimal tissue contact with the VLE probe, resulting in suboptimal in vivo visualization of the highly irregular glandular structures contained within the nodule. Routine use of this technology following resection, especially when images are compared with pathology specimens, can provide confirmation that the appropriate region of concern has been removed from the esophagus.