Influence of artificial intelligence on the diagnostic performance of endoscopists in the assessment of Barrett's esophagus: a tandem randomized and video trial.

BACKGROUND: This study evaluated the effect of an artificial intelligence (AI)-based clinical decision support system on the performance and diagnostic confidence of endoscopists in their assessment of Barrett's esophagus (BE). METHODS: 96 standardized endoscopy videos were assessed by 22 endoscopists with varying degrees of BE experience from 12 centers. Assessment was randomized into two video sets: group A (review first without AI and second with AI) and group B (review first with AI and second without AI). Endoscopists were required to evaluate each video for the presence of Barrett's esophagus-related neoplasia (BERN) and then decide on a spot for a targeted biopsy. After the second assessment, they were allowed to change their clinical decision and confidence level. RESULTS: AI had a stand-alone sensitivity, specificity, and accuracy of 92.2%, 68.9%, and 81.3%, respectively. Without AI, BE experts had an overall sensitivity, specificity, and accuracy of 83.3%, 58.1%, and 71.5%, respectively. With AI, BE nonexperts showed a significant improvement in sensitivity and specificity when videos were assessed a second time with AI (sensitivity 69.8% [95%CI 65.2%-74.2%] to 78.0% [95%CI 74.0%-82.0%]; specificity 67.3% [95%CI 62.5%-72.2%] to 72.7% [95%CI 68.2%-77.3%]). In addition, the diagnostic confidence of BE nonexperts improved significantly with AI. CONCLUSION: BE nonexperts benefitted significantly from additional AI. BE experts and nonexperts remained significantly below the stand-alone performance of AI, suggesting that there may be other factors influencing endoscopists' decisions to follow or discard AI advice.

Perspectives and awareness of endoscopy healthcare professionals on sustainable practices in gastrointestinal endoscopy: results of the LEAFGREEN survey.

BACKGROUND: Gastrointestinal (GI) endoscopy is one of healthcare's main contributors to climate change. We aimed to assess healthcare professionals' attitudes and the perceived barriers to implementation of sustainable GI endoscopy. METHODS: The LEAFGREEN web-based survey was a cross-sectional study conducted by the European Society of Gastrointestinal Endoscopy (ESGE) Green Endoscopy Working Group. The questionnaire comprised 39 questions divided into five sections (respondent demographics; climate change and sustainability beliefs; waste and resource management; single-use endoscopes and accessories; education and research). The survey was available via email to all active members of the ESGE and the European Society of Gastroenterology and Endoscopy Nurses and Associates (ESGENA) in March 2023. RESULTS: 407 respondents participated in the survey (11% response rate). Most participants (86%) agreed climate change is real and anthropogenic, but one-third did not consider GI endoscopy to be a significant contributor to climate change. Improvement in the appropriateness of endoscopic procedures (41%) and reduction in single-use accessories (34%) were considered the most important strategies to reduce the environmental impact of GI endoscopy. Respondents deemed lack of institutional support and knowledge from staff to be the main barriers to sustainable endoscopy. Strategies to reduce unnecessary GI endoscopic procedures and comparative studies of single-use versus reusable accessories were identified as research priorities. CONCLUSIONS: In this survey, ESGE and ESGENA members acknowledge climate change as a major threat to humanity. Further improvement in sustainability beliefs and professional attitudes, reduction in inappropriate GI endoscopy, and rational use of single-use accessories and endoscopes are critically required.

Diagnostic Value of Adding Magnifying Chromoendoscopy to Magnifying Narrow-Band Imaging Endoscopy for Colorectal Polyps.

BACKGROUND & AIMS: This study examined the additional value of magnifying chromoendoscopy (MCE) on magnifying narrow-band imaging endoscopy (M-NBI) in the optical diagnosis of colorectal polyps. METHODS: A multicenter prospective study was conducted at 9 facilities in Japan and Germany. Patients with colorectal polyps scheduled for resection were included. Optical diagnosis was performed by M-NBI first, followed by MCE. Both diagnoses were made in real time. MCE was performed on all type 2B lesions classified according to the Japan NBI Expert Team classification and other lesions at the discretion of endoscopists. The diagnostic accuracy and confidence of M-NBI and MCE for colorectal cancer (CRC) with deep invasion (≥T1b) were compared on the basis of histologic findings after resection. RESULTS: In total, 1173 lesions were included between February 2018 and December 2020, with 654 (5 hyperplastic polyp/sessile serrated lesion, 162 low-grade dysplasia, 403 high-grade dysplasia, 97 T1 CRCs, and 32 ≥T2 CRCs) examined using MCE after M-NBI. In the diagnostic accuracy for predicting CRC with deep invasion, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for M-NBI were 63.1%, 94.2%, 61.6%, 94.5%, and 90.2%, respectively, and for MCE they were 77.4%, 93.2%, 62.5%, 96.5%, and 91.1%, respectively. The sensitivity was significantly higher in MCE (P < .001). However, these additional values were limited to lesions with low confidence in M-NBI or the ones diagnosed as ≥T1b CRC by M-NBI. CONCLUSIONS: In this multicenter prospective study, we demonstrated the additional value of MCE on M-NBI. We suggest that additional MCE be recommended for lesions with low confidence or the ones diagnosed as ≥T1b CRC. Trials registry number: UMIN000031129.

Detection of duodenal villous atrophy on endoscopic images using a deep learning algorithm.

BACKGROUND AND AIMS: Celiac disease with its endoscopic manifestation of villous atrophy (VA) is underdiagnosed worldwide. The application of artificial intelligence (AI) for the macroscopic detection of VA at routine EGD may improve diagnostic performance. METHODS: A dataset of 858 endoscopic images of 182 patients with VA and 846 images from 323 patients with normal duodenal mucosa was collected and used to train a ResNet18 deep learning model to detect VA. An external dataset was used to test the algorithm, in addition to 6 fellows and 4 board-certified gastroenterologists. Fellows could consult the AI algorithm's result during the test. From their consultation distribution, a stratification of test images into "easy" and "difficult" was performed and used for classified performance measurement. RESULTS: External validation of the AI algorithm yielded values of 90%, 76%, and 84% for sensitivity, specificity, and accuracy, respectively. Fellows scored corresponding values of 63%, 72%, and 67% and experts scored 72%, 69%, and 71%, respectively. AI consultation significantly improved all trainee performance statistics. Although fellows and experts showed significantly lower performance for difficult images, the performance of the AI algorithm was stable. CONCLUSIONS: In this study, an AI algorithm outperformed endoscopy fellows and experts in the detection of VA on endoscopic still images. AI decision support significantly improved the performance of nonexpert endoscopists. The stable performance on difficult images suggests a further positive add-on effect in challenging cases.

One year of COVID-19 pandemic: Health care workers' infection rates and economical burden in medical facilities for oral and maxillofacial surgery.

The aim of this study was to create an overview on the COVID-associated burdens faced by the oral and maxillofacial surgery (OMS) workforce during 1 year of the pandemic. OMS hospitals and private practices nationwide were surveyed regarding health care worker (HCW) screening, infection status, pre-interventional testing, personal protective equipment (PPE), and economic impact. Participants were recruited via the German Society for Oral and Maxillofacial Surgery. A total of 11 hospitals (416 employees) and 55 private practices (744 employees) participated. The HCW infection rate was significantly higher in private practices than in clinics (4.7% vs. 1.4%, p<0.01), although most infections in HCW occurred in private environment (hospitals 88.2%, private practice 66.7%). Pre-interventional testing was performed significantly less for outpatients in private practices than in hospitals (90.7% vs. 36.4%, p<0.01). Polymerase chain reaction (PCR) was used significantly more for inpatients in hospitals than in private practices (100.0% vs. 27.3%, p<0.01). FFP2/3 use rose significantly in hospitals (0% in second quarter vs. 46% in fourth quarter, p<0.05) and private practices (15% in second quarter vs. 38% in fourth quarter, p<0.01). The decrease in procedures (≤50%) was significantly higher in hospitals than in private practices (90.9% vs. 40.0%, p<0.01). Despite higher infection rates in private practices, declining procedures and revenue affected hospitals more. Future COVID-related measures must adjust the infrastructure especially for hospitals to prevent further straining of staff and finances.

Vessel and tissue recognition during third-space endoscopy using a deep learning algorithm.

In this study, we aimed to develop an artificial intelligence clinical decision support solution to mitigate operator-dependent limitations during complex endoscopic procedures such as endoscopic submucosal dissection and peroral endoscopic myotomy, for example, bleeding and perforation. A DeepLabv3-based model was trained to delineate vessels, tissue structures and instruments on endoscopic still images from such procedures. The mean cross-validated Intersection over Union and Dice Score were 63% and 76%, respectively. Applied to standardised video clips from third-space endoscopic procedures, the algorithm showed a mean vessel detection rate of 85% with a false-positive rate of 0.75/min. These performance statistics suggest a potential clinical benefit for procedure safety, time and also training.

An artificial intelligence algorithm is highly accurate for detecting endoscopic features of eosinophilic esophagitis.

The endoscopic features associated with eosinophilic esophagitis (EoE) may be missed during routine endoscopy. We aimed to develop and evaluate an Artificial Intelligence (AI) algorithm for detecting and quantifying the endoscopic features of EoE in white light images, supplemented by the EoE Endoscopic Reference Score (EREFS). An AI algorithm (AI-EoE) was constructed and trained to differentiate between EoE and normal esophagus using endoscopic white light images extracted from the database of the University Hospital Augsburg. In addition to binary classification, a second algorithm was trained with specific auxiliary branches for each EREFS feature (AI-EoE-EREFS). The AI algorithms were evaluated on an external data set from the University of North Carolina, Chapel Hill (UNC), and compared with the performance of human endoscopists with varying levels of experience. The overall sensitivity, specificity, and accuracy of AI-EoE were 0.93 for all measures, while the AUC was 0.986. With additional auxiliary branches for the EREFS categories, the AI algorithm (AI-EoE-EREFS) performance improved to 0.96, 0.94, 0.95, and 0.992 for sensitivity, specificity, accuracy, and AUC, respectively. AI-EoE and AI-EoE-EREFS performed significantly better than endoscopy beginners and senior fellows on the same set of images. An AI algorithm can be trained to detect and quantify endoscopic features of EoE with excellent performance scores. The addition of the EREFS criteria improved the performance of the AI algorithm, which performed significantly better than endoscopists with a lower or medium experience level.

Antibody Response to SARS-CoV-2 Vaccination in Patients with Inflammatory Bowel Disease: Results of a Single-Center Cohort Study in a Tertiary Hospital in Germany.

BACKGROUND: COVID-19 is a viral disease caused by severe acute respiratory syndrome corona virus 2 (SARS-CoV-2), first described in 2019, with a significant impact on everyday life since then. In December 2020, the first vaccine against COVID-19 from BioNTech/Pfizer was approved for the first time. However, little is known about the immune response to vaccination in patients with inflammatory bowel disease (IBD) and immunomodulators or biologics. The aim of our study was to investigate antibody response to SARS-CoV-2 vaccination in patients with IBD receiving immunomodulators or biologics compared to healthy controls. METHODS: This was a single-center study with a retrospective observational design. Seventy-two patients with ulcerative colitis or Crohn's disease were included. Matching data from 72 healthy employees of our hospital were used as the control group. Data were matched by propensity score to patients with IBD. Blood samples were taken from both groups for antibody response, and both groups received an accompanying questionnaire. RESULTS: Sixty-five (90.3%) patients of the IBD group reported taking immunomodulatory therapy. The mean antibody level for all IBD patients was 1,257.1 U/mL (standard deviation [SD] 1,109.626) in males and 1,500.1 U/mL (SD 1142.760) in female IBD patients after full vaccination. Compared to the healthy group, reduced antibody response could be detected (IBD group 1,383.76 U/mL SD 1,125.617; control group 1,885.65 U/mL SD 727.572, p < 0.05). In this group, blood samples were taken with an average of 61.9 days after the first vaccination. There was no vaccination failure in the IBD group after 2 vaccinations. After the first vaccination, side effects, including muscle pain, pain at the injection site, and fatigue, were reported more often in IBD patients than in the control group (total symptoms IBD group 58.3%, control group 34.5%, p < 0.007). The opposite occurred after the second vaccination when side effects were higher in the control group (total symptoms IBD group 55.4%, control group 76%, p = 0.077). There was a trend to a reduced immune response in elderly patients. Disease duration and concomitant immunomodulatory therapy (TNF-alpha blockers, interleukin inhibitors, integrin inhibitors, methotrexate, or azathioprine) had no impact on the immune response. However, longer time to last medication given and time passed to vaccination in patients with IBD seems to have a positive impact on antibody levels. CONCLUSION: Overall, we could show a high antibody response to vaccination with COVID-19 in all patients with IBD after 2 vaccinations. Vaccination was well tolerated, and no other adverse events were detected. Concomitant immunomodulatory therapy (TNF-alpha blockers, interleukin inhibitors, integrin inhibitors, methotrexate, or azathioprine) had no impact on seroconversion. Further evaluation of antibody titers over time is mandatory to detect early the need for re-vaccination in these patients.

[Impact of Covid 19 on endoscopy in Germany]. / Ein Jahr Covid-19: Testung, Verwendung von Schutzausrüstung und Auswirkungen auf die Gastrointestinale Endoskopie in Deutschland.

BACKGROUND: Practices and hospitals are facing great challenges in coping with the COVID-19-pandemic. So far, data on the impact of the pandemic on gastroenterological facilities are lacking, especially on a temporal course. A database is lacking, especially for the outpatient care sector. University Hospital of Augsburg was commissioned to generate data on this as a part of the collaborative project B-FAST of the Network of University Medicine (NUM). METHODS: Gastroenterological institutions nationwide were surveyed by an online questionnaire. Recruitment was carried out via the German Society of Gastroenterology, Digestive and Metabolic Diseases (DGVS) and the Professional Association of Gastroenterologists in Private Practice (bng). This manuscript provides an overview of data on the use of protective equipment, pre-interventional testing of patients, staff screening and economic impact over the course of the pandemic. RESULTS: 429 facilities answered the questionnaire. Practices tested their patients pre-interventionally significantly less often than clinics (7.8% vs. 82.6%). In clinics, inpatients (93.1%) were tested significantly more often than outpatients (72.2%). The use of personal protective equipment (PPE) increased significantly during the pandemic. It was shown that over 70% of facilities screened their staff for SARS-CoV-2 without cause. Clinics cancelled elective procedures significantly more often than practices in quarter 4/2020. Procedures and turnover decreased in 2020 compared to the previous year. However, fewer facilities were affected by a loss of revenue than expected in previous studies. CONCLUSION: Our data demonstrate the variable implementation of pre-interventional SARS-CoV-2 testing in outpatient and inpatient care. The use of adequate PPE and staff screening increased during the pandemic.

Gastrointestinal bleeding and endoscopic findings in critically and non-critically ill patients with corona virus disease 2019 (COVID-19): Results from Lean European Open Survey on SARS-CoV-2 (LEOSS) and COKA registries.

BACKGROUND: Corona virus disease 2019 (COVID-19) patients are at increased risk for thromboembolic events. It is unclear whether the risk for gastrointestinal (GI) bleeding is also increased. METHODS: We considered 4128 COVID-19 patients enrolled in the Lean European Open Survey on SARS-CoV-2 (LEOSS) registry. The association between occurrence of GI bleeding and comorbidities as well as medication were examined. In addition, 1216 patients from COKA registry were analyzed focusing on endoscopy diagnostic findings. RESULTS: A cumulative number of 97 patients (1.8%) with GI bleeding were identified in the LEOSS registry and COKA registry. Of 4128 patients from the LEOSS registry, 66 patients (1.6%) had a GI bleeding. The rate of GI bleeding in patients with intensive care unit (ICU) admission was 4.5%. The use of therapeutic dose of anticoagulants showed a significant association with the increased incidence of bleeding in the critical phase of disease. The Charlson comorbidity index and the COVID-19 severity index were significantly higher in the group of patients with GI bleeding than in the group of patients without GI bleeding (5.83 (SD = 2.93) vs. 3.66 (SD = 3.06), p < 0.01 and 3.26 (SD = 1.69) vs. 2.33 (SD = 1.53), p < 0.01, respectively). In the COKA registry 31 patients (2.5%) developed a GI bleeding. Of these, the source of bleeding was identified in upper GI tract in 21 patients (67.7%) with ulcer as the most frequent bleeding source (25.8%, n = 8) followed by gastroesophageal reflux (16.1%, n = 5). In three patients (9.7%) GI bleeding source was located in lower GI tract caused mainly by diverticular bleeding (6.5%, n = 2). In seven patients (22.6%) the bleeding localization remained unknown. CONCLUSION: Consistent with previous research, comorbidities and disease severity correlate with the incidence of GI bleeding. Also, therapeutic anticoagulation seems to be associated with a higher risk of GI bleeding. Overall, the risk of GI bleeding seems not to be increased in COVID-19 patients.

[Screening gastroscopy in Germany - first results of the pilot phase of the Prä-GIT study]. / Vorsorgegastroskopie in Deutschland ­ erste Ergebnisse der Pilotphase der Prä-GIT-Studie.

The primary aim of the Prä-GIT study is to collect information on the prevalence and course of precancerous and early malignant lesions in the upper GIT (uGIT) as well as risk factors associated with these lesions in asymptomatic individuals. Study participants presenting for a screening colonoscopy will undergo an additional endoscopic examination of the uGIT in the same session. 5000 participants in 30 endoscopy clinics in Bayern will be included. The pilot study presented here was performed to test the main study protocol as well as the acceptance of an additional uGIT endoscopic examination. METHODS: Three endoscopy clinics in Bayern took part in the three-month pilot study between October and December 2018. Patients presenting for a screening colonoscopy in these clinics were offered an additional endoscopic examination of the uGIT which was performed according to a standard operating procedure (SOP). Furthermore, data on dietary habits and lifestyle, as well as biological samples, were collected. RESULTS: 52 participants were included in three clinics in Altötting, Augsburg and Regensburg. The average age was 63.4 years. The average time spent per uGIT endoscopy was 11 minutes. No complications occurred. One participant showed a polyp of the recessus piriformis. Refluxesophagitis was seen in 21 participants, four participants had Barrett's esophagus with histological evidence of low-grade dysplasia in one patient, and an esophageal polyp with low-grade dysplasia was seen in one participant. Helicobacter-pylori gastritis was documented in 12 participants. Corpus-dominant atrophy was described in one participant, while a duodenal adenoma with low-grade dysplasia was seen in two participants. 100 % and 89 % of study participants gave a blood or a stool sample, respectively. All participants answered the questionnaire on dietary and lifestyle habits. CONCLUSION: The Prä-GIT study was designed to generate a database for premalignant and early cancerous lesions of the uGIT in Germany in asymptomatic individuals presenting for a regular screening colonoscopy. The current pilot study has delivered important insights into the feasibility of the main study protocol. Additionally, the pilot study has shown a high rate of acceptance of an additional uGIT examination in this study population.

Endoscopic Upper GI Screening.

The rationale behind endoscopic screening of the upper gastrointestinal tract (GIT) in the West is different from Asian countries with a high prevalence of gastric cancer. If epidemiological data on upper GIT cancer in the Western world is considered, then endoscopic screening of the upper GIT must target premalignant conditions as well and, furthermore, lay emphasis on the assessment of the entire upper GIT in high-risk individuals from the mouth till the duodenum. Although data on the incidence, prevalence, and further development of premalignant conditions in the upper GIT is lacking, a risk-based approach to endoscopic screening will be more cost-effective than a general screening approach. Finally, endoscopists need to be educated and trained in the assessment of the upper GIT during screening endoscopy.

Step-up training for colorectal and gastric ESD and the challenge of ESD training in the proximal colon: results from a German Center.

BACKGROUND AND STUDY AIMS: The endoscopic submucosal dissection (ESD) of cancerous and precancerous lesions in the gastrointestinal (GI) tract is an effective but difficult-to-learn procedure, especially for Western endoscopists. We developed a step-up training protocol and evaluated its outcome on a single endoscopist using specific performance measures. PATIENTS AND METHODS: The training protocol included a 12-month period of 50 observational and supervised ESD cases as well as 24 animal procedures. After completion of the protocol, ESD competency was evaluated in the first 30 independent ESD cases. Majority of lesions were located in the stomach and rectum. Performance measures included R0-resection rate, complication rate and resection speed. RESULTS: R0 resection rate was 93 %, complication rate was 7 % and median resection speed was 6,77 cm 2 /h. 1 case of delayed perforation with subsequent hemicolectomy occurred in the descending colon. CONCLUSION: The ESD step-up training protocol used in this study on a single endoscopist showed excellent outcomes for lesions in the rectum and stomach. However, ESD training in the upper colon remains a challenge.

Endogenous and exogenous pentraxin-3 limits postischemic acute and chronic kidney injury.

Ischemia-reperfusion activates innate immunity and sterile inflammation, resulting in acute kidney injury. Since pentraxin 3 (PTX3) regulates multiple aspects of innate immunity and tissue inflammation, we tested whether PTX3 would be involved in renal ischemia-reperfusion injury. Renal pedicle clamping increased PTX3 serum levels, as well as PTX3 expression, inside the kidney but predominantly in CD45/CD11c(+) cells, a subpopulation of intrarenal mononuclear phagocytes. Lack of PTX3 aggravated postischemic acute kidney injury as evidenced by massive tubular necrosis, and TNF and IL-6 release, as well as massively increased neutrophil and macrophage infiltrates at 24 h. This was followed by tubular atrophy, interstitial fibrosis, and kidney shrinking 10 weeks later. In vivo microscopy uncovered increased leukocyte adhesion and transmigration in postischemic microvessels of Ptx3-deficient mice. Furthermore, injection of recombinant PTX3 up to 6 h after reperfusion prevented renal leukocyte recruitment and postischemic kidney injury. Thus, local PTX3 release from a subpopulation of intrarenal mononuclear phagocytes or delayed PTX3 treatment limits postischemic renal inflammation. Conversely, Ptx3 loss-of-function mutations predispose to postischemic acute kidney injury and subsequent chronic kidney disease.

Pentraxins in nephrology: C-reactive protein, serum amyloid P and pentraxin-3.

Every clinician uses C-reactive protein (CRP) levels as a biomarker for systemic inflammation in acute disorders. Nephrologists also consider CRP levels as a predictor for overall mortality in patients with chronic kidney disease or end-stage renal disease. But what is the biological function of CRP? CRP is a member of the family of pentraxins, which are small pentameric innate immunity effector proteins. Pentraxins are absent or weakly expressed during homeostasis. However, the pro-inflammatory cytokines interleukin (IL)-1, IL-6 and tumour necrosis factor induce CRP and serum amyloid P (SAP) in hepatocytes, whereas the long pentraxins, such as pentraxin (PTX)-3, are produced in peripheral tissues and monocytic phagocytes. Pentraxins opsonize pathogens or other particles such as dead cells, for their phagocytic clearance or induce pathogen killing in extracellular compartments. In this review, we discuss the immunoregulatory properties of the different members of the pentraxin family. We discuss the evolving evidence demonstrating their roles in acute and chronic forms of kidney disease and the significance of SAP and PTX3 as additional biomarkers of innate immune activation and systemic inflammation.