Search

Background: Colorectal cancer is a leading cause of cancer-related deaths worldwide. The best method to prevent CRC is a colonoscopy. However, not all colon polyps have the risk of becoming cancerous. Therefore, polyps are classified using different classification systems. After the classification, further treatment and procedures are based on the classification of the polyp. Nevertheless, classification is not easy. Therefore, we suggest two novel automated classifications system assisting gastroenterologists in classifying polyps based on the NICE and Paris classification., Methods: We build two classification systems. One is classifying polyps based on their shape (Paris). The other classifies polyps based on their texture and surface patterns (NICE). A two-step process for the Paris classification is introduced: First, detecting and cropping the polyp on the image, and secondly, classifying the polyp based on the cropped area with a transformer network. For the NICE classification, we design a few-shot learning algorithm based on the Deep Metric Learning approach. The algorithm creates an embedding space for polyps, which allows classification from a few examples to account for the data scarcity of NICE annotated images in our database., Results: For the Paris classification, we achieve an accuracy of 89.35 %, surpassing all papers in the literature and establishing a new state-of-the-art and baseline accuracy for other publications on a public data set. For the NICE classification, we achieve a competitive accuracy of 81.13 % and demonstrate thereby the viability of the few-shot learning paradigm in polyp classification in data-scarce environments. Additionally, we show different ablations of the algorithms. Finally, we further elaborate on the explainability of the system by showing heat maps of the neural network explaining neural activations., Conclusion: Overall we introduce two polyp classification systems to assist gastroenterologists. We achieve state-of-the-art performance in the Paris classification and demonstrate the viability of the few-shot learning paradigm in the NICE classification, addressing the prevalent data scarcity issues faced in medical machine learning., (© 2023. The Author(s).)

Background: Secondary sclerosing cholangitis (SSC) is a rare disease with poor prognosis. Cases of SSC have been reported following coronavirus disease 2019 (COVID-SSC). The aim of this study was to compare COVID-SSC to SSC in critically ill patients (SSC-CIP) and to assess factors influencing transplant-free survival., Methods: In this retrospective, multicenter study involving 127 patients with SSC from 9 tertiary care centers in Germany, COVID-SSC was compared to SSC-CIP and logistic regression analyses were performed investigating factors impacting transplant-free survival., Results: Twenty-four patients had COVID-SSC, 77 patients SSC-CIP, and 26 patients other forms of SSC. COVID-SSC developed after a median of 91 days following COVID-19 diagnosis. All patients had received extensive intensive care treatment (median days of mechanical ventilation, 48). Patients with COVID-SSC and SSC-CIP were comparable in most of the clinical parameters and transplant-free survival was not different from other forms of SSC (P = .443, log-rank test). In the overall cohort, the use of ursodeoxycholic acid (UDCA) (odds ratio [OR], 0.36 [95% confidence interval {CI}, .16-.80], P = .013; log-rank P < .001) and high serum albumin levels (OR, 0.40 [95% CI, .17-.96], P = .040) were independently associated with an increased transplant-free survival, while the presence of liver cirrhosis (OR, 2.52 [95% CI, 1.01-6.25], P = .047) was associated with worse outcome. Multidrug-resistant organism (MDRO) colonization or infection did not impact patients' survival., Conclusions: COVID-SSC and CIP-SSC share the same clinical phenotype, course of the disease, and risk factors for its development. UDCA may be a promising therapeutic option in SSC, though future prospective trials are needed to confirm our findings., Competing Interests: Potential conflicts of interest. D. B. has served as a consultant for Bayer Healthcare, Boston Scientific, and Shionogi, and has given lectures for the Falk Foundation. K. H. has received an unrestricted research grant from Grifols; speaker’s fees from Grifols, CSL Behring, AbbVie, and Chiesi; and travel support from AbbVie. K. H. has also received support for the present manuscript from the START program within the medical faculty at Rheinisch-Westfälische Technische Hochschule Aachen University, the ALTA Award from Grifols, and the German Liver Foundation. V. T. M. has received travel support from AbbVie. F. F. has received travel support from AbbVie and Novartis, and speaker’s fees from AbbVie, MSD, Ipsen, and Fresenius. K. H. P. has received payment or honoraria from AbbVie and Gilead and support for attending meetings and/or travel from AbbVie. S. Z. has received speaking and/or consulting fees from AbbVie, Allergan, BioMarin, Bristol-Myers Squibb (BMS), Falk, Gilead, Intercept, Janssen, Novo Nordisk, Sobi, Theratechnologies, and Merck/MSD, and payment for expert testimony from Gilead. O. W. has received fees for advisory board membership from Amgen, Bayer, BMS, Celgene, Eisai, Incyte, Ipsen, Merck Serono, MSD, Novartis, Roche, Servier, and Shire; spearker’s fees from AstraZeneca, Bayer, BMS, Eisai, Ipsen, MSD, Novartis, Roche, and Shire; travel support from AbbVie, Bayer, BMS, Gilead, Ipsen, Medac, and Merck Serono; and funding for investigator-initiated trials from Else Kröner-Fresenius-Stiftung, Medac, and Merck Serono. He is also an investigator for Basilea, Incyte, and MSD. F. K. has received payment or honoraria from Eisai, Sirtex, and Ipsen, and support for attending meetings and/or travel from Janssen and Pfizer. M. M. M. has received speaker’s fees from AbbVie; travel support from AbbVie, Gilead, and Intercept; research grant from Gilead; and consulting fees and participation on a data and safety monitoring board or advisory board from Sobi. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed., (© The Author(s) 2022. Published by Oxford University Press on behalf of Infectious Diseases Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

Colorectal cancer (CRC) is a leading cause of cancer-related deaths worldwide. The best method to prevent CRC is with a colonoscopy. During this procedure, the gastroenterologist searches for polyps. However, there is a potential risk of polyps being missed by the gastroenterologist. Automated detection of polyps helps to assist the gastroenterologist during a colonoscopy. There are already publications examining the problem of polyp detection in the literature. Nevertheless, most of these systems are only used in the research context and are not implemented for clinical application. Therefore, we introduce the first fully open-source automated polyp-detection system scoring best on current benchmark data and implementing it ready for clinical application. To create the polyp-detection system (ENDOMIND-Advanced), we combined our own collected data from different hospitals and practices in Germany with open-source datasets to create a dataset with over 500,000 annotated images. ENDOMIND-Advanced leverages a post-processing technique based on video detection to work in real-time with a stream of images. It is integrated into a prototype ready for application in clinical interventions. We achieve better performance compared to the best system in the literature and score a F1-score of 90.24% on the open-source CVC-VideoClinicDB benchmark.

Background and Aims: Computer-aided polyp detection (CADe) may become a standard for polyp detection during colonoscopy. Several systems are already commercially available. We report on a video-based benchmark technique for the first preclinical assessment of such systems before comparative randomized trials are to be undertaken. Additionally, we compare a commercially available CADe system with our newly developed one., Methods: ENDOTEST consisted in the combination of two datasets. The validation dataset contained 48 video-snippets with 22,856 manually annotated images of which 53.2% contained polyps. The performance dataset contained 10 full-length screening colonoscopies with 230,898 manually annotated images of which 15.8% contained a polyp. Assessment parameters were accuracy for polyp detection and time delay to first polyp detection after polyp appearance (FDT). Two CADe systems were assessed: a commercial CADe system (GI-Genius, Medtronic), and a self-developed new system (ENDOMIND). The latter being a convolutional neuronal network trained on 194,983 manually labeled images extracted from colonoscopy videos recorded in mainly six different gastroenterologic practices., Results: On the ENDOTEST, both CADe systems detected all polyps in at least one image. The per-frame sensitivity and specificity in full colonoscopies was 48.1% and 93.7%, respectively for GI-Genius; and 54% and 92.7%, respectively for ENDOMIND. Median FDT of ENDOMIND with 217 ms (Inter-Quartile Range(IQR)8-1533) was significantly faster than GI-Genius with 1050 ms (IQR 358-2767, p  = 0.003)., Conclusions: Our benchmark ENDOTEST may be helpful for preclinical testing of new CADe devices. There seems to be a correlation between a shorter FDT with a higher sensitivity and a lower specificity for polyp detection.

Background: Multiple computer-aided systems for polyp detection (CADe) have been introduced into clinical practice, with an unclear effect on examiner behavior. This study aimed to measure the influence of a CADe system on reaction time, mucosa misinterpretation, and changes in visual gaze pattern., Methods: Participants with variable levels of colonoscopy experience viewed video sequences (n = 29) while eye movement was tracked. Using a crossover design, videos were presented in two assessments, with and without CADe support. Reaction time for polyp detection and eye-tracking metrics were evaluated., Results: 21 participants performed 1218 experiments. CADe was significantly faster in detecting polyps compared with participants (median 1.16 seconds [99 %CI 0.40-3.43] vs. 2.97 seconds [99 %CI 2.53-3.77], respectively). However, the reaction time of participants when using CADe (median 2.90 seconds [99 %CI 2.55-3.38]) was similar to that without CADe. CADe increased misinterpretation of normal mucosa and reduced the eye travel distance., Conclusions: Results confirm that CADe systems detect polyps faster than humans. However, use of CADe did not improve human reaction times. It increased misinterpretation of normal mucosa and decreased the eye travel distance. Possible consequences of these findings might be prolonged examination time and deskilling., Competing Interests: The authors declare that they have no conflict of interest., (The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/licenses/by-nc-nd/4.0/).)

Purpose:  Hepatic hydrothorax (HH) is defined as transudate in the pleural cavity in patients with decompensated liver cirrhosis (DC) without concomitant cardiopulmonary or pleural disease. It is associated with high short-term mortality. HH can evolve via translocation through diaphragmatic gaps. The aim of this study was to evaluate the feasibility and safety of injecting ultrasound contrast medium into the peritoneal cavity to detect HH., Materials and Methods:  This study included patients with concomitant ascites and pleural effusion who were admitted to our hospital between March 2009 and February 2019. A peritoneal catheter was inserted and ultrasound contrast medium was injected into the peritoneal cavity. In parallel, the peritoneal and pleural cavities were monitored for up to 10 minutes., Results:  Overall, 43 patients were included. The median age was 60 years and the majority of patients were male (n = 32, 74 %). Most patients presented with right-sided pleural effusion (n = 32, 74 %), 3 (7 %) patients with left-sided and 8 (19 %) patients had bilateral pleural effusion. In 12 (28 %) patients ascites puncture was not safe due to low volume ascites. Thus, the procedure could be performed in 31 (72 %) patients. No adverse events occurred. In 16 of 31 (52 %) patients we could visualize a trans-diaphragmic flow of microbubbles. The median time until transition was 120 seconds., Conclusion:  Our clinical real-world experience supports the safety and feasibility of intraperitoneal ultrasound contrast medium application to detect HH in patients with DC, as a non-radioactive real-time visualization of HH. Our study comprises the largest cohort and longest experience using this method to date., Competing Interests: The authors declare that they have no conflict of interest., (Thieme. All rights reserved.)

Purpose: Computer-aided polyp detection (CADe) systems for colonoscopy are already presented to increase adenoma detection rate (ADR) in randomized clinical trials. Those commercially available closed systems often do not allow for data collection and algorithm optimization, for example regarding the usage of different endoscopy processors. Here, we present the first clinical experiences of a, for research purposes publicly available, CADe system., Methods: We developed an end-to-end data acquisition and polyp detection system named EndoMind. Examiners of four centers utilizing four different endoscopy processors used EndoMind during their clinical routine. Detected polyps, ADR, time to first detection of a polyp (TFD), and system usability were evaluated (NCT05006092)., Results: During 41 colonoscopies, EndoMind detected 29 of 29 adenomas in 66 of 66 polyps resulting in an ADR of 41.5%. Median TFD was 130 ms (95%-CI, 80-200 ms) while maintaining a median false positive rate of 2.2% (95%-CI, 1.7-2.8%). The four participating centers rated the system using the System Usability Scale with a median of 96.3 (95%-CI, 70-100)., Conclusion: EndoMind's ability to acquire data, detect polyps in real-time, and high usability score indicate substantial practical value for research and clinical practice. Still, clinical benefit, measured by ADR, has to be determined in a prospective randomized controlled trial., (© 2022. The Author(s).)

Background: Machine learning, especially deep learning, is becoming more and more relevant in research and development in the medical domain. For all the supervised deep learning applications, data is the most critical factor in securing successful implementation and sustaining the progress of the machine learning model. Especially gastroenterological data, which often involves endoscopic videos, are cumbersome to annotate. Domain experts are needed to interpret and annotate the videos. To support those domain experts, we generated a framework. With this framework, instead of annotating every frame in the video sequence, experts are just performing key annotations at the beginning and the end of sequences with pathologies, e.g., visible polyps. Subsequently, non-expert annotators supported by machine learning add the missing annotations for the frames in-between., Methods: In our framework, an expert reviews the video and annotates a few video frames to verify the object's annotations for the non-expert. In a second step, a non-expert has visual confirmation of the given object and can annotate all following and preceding frames with AI assistance. After the expert has finished, relevant frames will be selected and passed on to an AI model. This information allows the AI model to detect and mark the desired object on all following and preceding frames with an annotation. Therefore, the non-expert can adjust and modify the AI predictions and export the results, which can then be used to train the AI model., Results: Using this framework, we were able to reduce workload of domain experts on average by a factor of 20 on our data. This is primarily due to the structure of the framework, which is designed to minimize the workload of the domain expert. Pairing this framework with a state-of-the-art semi-automated AI model enhances the annotation speed further. Through a prospective study with 10 participants, we show that semi-automated annotation using our tool doubles the annotation speed of non-expert annotators compared to a well-known state-of-the-art annotation tool., Conclusion: In summary, we introduce a framework for fast expert annotation for gastroenterologists, which reduces the workload of the domain expert considerably while maintaining a very high annotation quality. The framework incorporates a semi-automated annotation system utilizing trained object detection models. The software and framework are open-source., (© 2022. The Author(s).)

Competing Interests: Die Autorinnen/Autoren geben an, dass kein Interessenkonflikt besteht.

Background: Artificial intelligence (AI) using deep learning methods for polyp detection (CADe) and characterization (CADx) is on the verge of clinical application. CADe already implied its potential use in randomized controlled trials. Further efforts are needed to take CADx to the next level of development., Aim: This work aims to give an overview of the current status of AI in colonoscopy, without going into too much technical detail., Methods: A literature search to identify important studies exploring the use of AI in colonoscopy was performed., Results: This review focuses on AI performance in screening colonoscopy summarizing the first prospective trials for CADe, the state of research in CADx as well as current limitations of those systems and legal issues., (© 2021 The Authors. United European Gastroenterology Journal published by Wiley Periodicals LLC. on behalf of United European Gastroenterology.)

Background: The use of an indwelling peritoneal catheter system in hospitalized patients with ascites could facilitate patient management by the prevention of repetitive abdominal paracentesis. Despite these possible benefits, the use of indwelling catheters is not widely established., Objective: This retrospective study aimed to evaluate the feasibility, effectiveness and safety of the use of an indwelling catheter for ascites drainage in the clinical routine., Methods: This retrospective study included all indwelling peritoneal catheter placements in our department in hospitalized patients with cirrhosis between 2014 and 2017., Results: A total of 324 indwelling catheter placements for ascites in 192 hospitalized patients with cirrhosis were included. The catheter (7F, 8 cm) was placed ultrasound-assisted bed-side on the hospital ward. The technical success rate of the catheter placement was 99.7% (323/324). In 17.5% (64/324) the catheter was placed to optimize ascitic drainage prior to an abdominal intervention (e.g. transjugular intrahepatic portosystemic shunt). The median time of catheter retention was 48 hours (8-168 hours) and the median cumulative amount of drained ascites 8000 ml (550-28,000). The most common adverse event was acute kidney injury (49/324, 15.1%); the risk was particularly higher in patients with a Model for End-Stage Liver Disease (MELD) score ≥ 16 ( p  = 0.028; odds ratio 2.039). Ascitic fistula after catheter removal was observed in 9.6% (31/324). Catheter-related infections occurred in 4.3% (14/324), and bleeding was documented in three cases (0.8%) with one major bleeding (0.3%)., Conclusion: The placement of an indwelling catheter for repetitive ascitic drainage in hospitalized patients with cirrhosis can be established in the clinical routine, facilitating patient management. High-MELD patients especially have to be monitored for acute kidney injury.

Interleukin-22 (IL-22) has context-dependent hepatoprotective or adverse properties in vitro and in animal models. IL-22 binding protein (IL-22BP) is a soluble inhibitor of IL-22 signaling. The role of IL-22 and IL-22BP in patients with acute-on-chronic liver failure (ACLF) is unclear. Beginning in August 2013, patients with liver cirrhosis with and without ACLF were prospectively enrolled and followed at predefined time points. IL-22 and IL-22BP concentrations were quantified and associated with clinical endpoints. The impact of IL-22BP on hepatocellular IL-22 signaling was assessed by functional experiments. A total of 139 patients were analyzed, including 45 (32%), 52 (37%), and 42 (30%) patients with compensated/stable decompensated liver cirrhosis, acute decompensation of liver cirrhosis, and ACLF at baseline, respectively. Serum levels of IL-22 and IL-22BP were strongly associated with the presence of, or progression to, ACLF ( P < 0.001), and with mortality ( P < 0.01). Importantly, the mean IL-22BP levels exceeded IL-22 levels more than 300-fold. Furthermore, IL-22BP/IL-22 ratios were lowest in patients with adverse outcomes (i.e., ACLF and death). In vitro experiments showed that IL-22BP at these concentrations inhibits hepatocellular IL-22 signaling, including the induction of acute-phase proteins. The capacity of patient serum to induce signal transducer and activator of transcription 3 phosphorylation was substantially higher in the presence of low versus high IL-22BP/IL-22 ratios. Conclusion: Our study reveals that high IL-22 levels and low ratios of IL-22BP/IL-22 are associated with ACLF and mortality of patients with cirrhosis. Excessive secretion of IL-22BP can neutralize IL-22 in vitro and may prevent-likely in a context-specific manner-hepatoprotective, but also adverse effects, of IL-22 in patients with cirrhosis.

Background and Aim: This is a prospective study for evaluation of 2D-shear wave elastography (2D-SWE) for characterisation and differentiation of benign und malignant focal liver lesions (FLLs)., Methods: The patients referred to our ultrasound unit were prospectively included. B-mode ultrasound and 2D-SWE (Aixplorer® France) were performed for one FLL in each patient. Liver histology and/or contrast-enhanced imaging were used as a reference method., Results: 140 patients with FLL were included. SWE acquisitions failed in 24% of them. Therefore, 106 patients with FLL could be analysed, 42/106 with benign and 64/106 with malignant FLLs. The median stiffness for benign FLLs was 16.4 (2.1-71.9) kPa: 16.55 kPa for 18 focal nodular hyperplasia (FNH), 16.35 kPa for 18 hemangioma, 9.8 kPa for 3 focal fatty sparings (FFS), 8.9 kPa for 1 adenoma, 20 kPa for one regenerative node and 29 kPa for one cholangiofibroma, and for the malignant FLLs 36 (4.1-142.9) kPa: 44.8 kPa for 16 hepatocellular carcinoma (HCC), 70.7 kPa for 7 cholangiocarcinoma (CCC) and 29.5 kPa for the 41 metastasis (p<0.001). Malignant FLLs were significantly stiffer than benign FLLs (p<0.0001). Cholangiocarcinomas were the stiffest malignant FFLs with significantly higher values as compared to HCCs and metastases (p=0.033 and p=0.0079, respectively). No significant difference in stiffness could be observed between the different benign FLL entities. No significant difference was observed whether 2D-SWE included the whole FLL, the periphery or only the hardest area of the FLL., Conclusions: 2D-SWE provides further characterising information for interpretation of FLLs and may be useful at least in differentiation of CCCs and HCCs.

Two-dimensional shear wave elastography (2-D SWE) is an ultrasound-based elastography method integrated into a conventional ultrasound machine. It can evaluate larger regions of interest and, therefore, might be better at determining the overall fibrosis distribution. The aim of this prospective study was to compare 2-D SWE with the two best evaluated liver elastography methods, transient elastography and acoustic radiation force impulse (point SWE using acoustic radiation force impulse) imaging, in the same population group. The study included 132 patients with chronic hepatopathies, in which liver stiffness was evaluated using transient elastography, acoustic radiation force impulse imaging and 2-D SWE. The reference methods were liver biopsy for the assessment of liver fibrosis (n = 101) and magnetic resonance imaging/computed tomography for the diagnosis of liver cirrhosis (n = 31). No significant difference in diagnostic accuracy, assessed as the area under the receiver operating characteristic curve (AUROC), was found between the three elastography methods (2-D SWE, transient elastography, acoustic radiation force impulse imaging) for the diagnosis of significant and advanced fibrosis and liver cirrhosis in the "per protocol" (AUROCs for fibrosis stages ≥2: 0.90, 0.95 and 0.91; for fibrosis stage [F] ≥3: 0.93, 0.95 and 0.94; for F = 4: 0.92, 0.96 and 0.92) and "intention to diagnose" cohort (AUROCs for F ≥2: 0.87, 0.92 and 0.91; for F ≥3: 0.91, 0.93 and 0.94; for F = 4: 0.88, 0.90 and 0.89). Therefore, 2-D SWE, ARFI imaging and transient elastography seem to be comparably good methods for non-invasive assessment of liver fibrosis., (Copyright © 2015 World Federation for Ultrasound in Medicine & Biology. Published by Elsevier Inc. All rights reserved.)