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.

Evaluation of a single-use gastroscope in patients presenting with suspected upper gastrointestinal hemorrhage: a pilot feasibility study (One-Scope I).

BACKGROUND : Outbreaks of multidrug-resistant bacteria due to contaminated duodenoscopes and infection risks during the COVID-19 pandemic have driven the development of single-use endoscopes. The first single-use gastroscope is now available in Europe. Besides waste disposal and cost issues, the infection risk and performance remain unclear. We aimed to evaluate a single-use gastroscope in patients with signs of upper gastrointestinal bleeding. METHODS : 20 consecutive patients presenting with clinical signs of upper gastrointestinal bleeding between October and November 2022 were included in this case series. The primary aim was technical success, defined as access to the descending duodenum and adequate assessment of the upper gastrointestinal tract for the presence of a bleeding site. RESULTS : The primary aim was achieved in 19/20 patients (95 %). The bleeding site was identified in 18 patients. A therapeutic intervention was performed in six patients (two cap-mounted clips, one standard hemostatic clip, two variceal band ligations, one hemostatic powder, two adrenaline injections); technical and clinical success were achieved in all six patients. Two crossovers to a standard gastroscope occurred. CONCLUSIONS : Use of single-use gastroscopes may be feasible for patients presenting for urgent endoscopic evaluation and treatment of upper gastrointestinal bleeding.

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.

Endoscopic prediction of submucosal invasion in Barrett's cancer with the use of artificial intelligence: a pilot study.

BACKGROUND: The accurate differentiation between T1a and T1b Barrett's-related cancer has both therapeutic and prognostic implications but is challenging even for experienced physicians. We trained an artificial intelligence (AI) system on the basis of deep artificial neural networks (deep learning) to differentiate between T1a and T1b Barrett's cancer on white-light images. METHODS: Endoscopic images from three tertiary care centers in Germany were collected retrospectively. A deep learning system was trained and tested using the principles of cross validation. A total of 230 white-light endoscopic images (108 T1a and 122 T1b) were evaluated using the AI system. For comparison, the images were also classified by experts specialized in endoscopic diagnosis and treatment of Barrett's cancer. RESULTS: The sensitivity, specificity, F1 score, and accuracy of the AI system in the differentiation between T1a and T1b cancer lesions was 0.77, 0.64, 0.74, and 0.71, respectively. There was no statistically significant difference between the performance of the AI system and that of experts, who showed sensitivity, specificity, F1, and accuracy of 0.63, 0.78, 0.67, and 0.70, respectively. CONCLUSION: This pilot study demonstrates the first multicenter application of an AI-based system in the prediction of submucosal invasion in endoscopic images of Barrett's cancer. AI scored equally to international experts in the field, but more work is necessary to improve the system and apply it to video sequences and real-life settings. Nevertheless, the correct prediction of submucosal invasion in Barrett's cancer remains challenging for both experts and AI.

Conventional transarterial chemoembolization versus drug-eluting bead transarterial chemoembolization for the treatment of hepatocellular carcinoma.

BACKGROUND: To compare the overall survival of patients with hepatocellular carcinoma (HCC) who were treated with lipiodol-based conventional transarterial chemoembolization (cTACE) with that of patients treated with drug-eluting bead transarterial chemoembolization (DEB-TACE). METHODS: By an electronic search of our radiology information system, we identified 674 patients that received TACE between November 2002 and July 2013. A total of 520 patients received cTACE, and 154 received DEB-TACE. In total, 424 patients were excluded for the following reasons: tumor type other than HCC (n=91), liver transplantation after TACE (n=119), lack of histological grading (n=58), incomplete laboratory values (n=15), other reasons (e.g., previous systemic chemotherapy) (n=114), or were lost to follow-up (n=27). Therefore, 250 patients were finally included for comparative analysis (n=174 cTACE; n=76 DEB-TACE). RESULTS: There were no significant differences between the two groups regarding sex, overall status (Barcelona Clinic Liver Cancer classification), liver function (Child-Pugh), portal invasion, tumor load, or tumor grading (all p>0.05). The mean number of treatment sessions was 4±3.1 in the cTACE group versus 2.9±1.8 in the DEB-TACE group (p=0.01). Median survival was 409 days (95% CI: 321-488 days) in the cTACE group, compared with 369 days (95% CI: 310-589 days) in the DEB-TACE group (p=0.76). In the subgroup of Child A patients, the survival was 602 days (484-792 days) for cTACE versus 627 days (364-788 days) for DEB-TACE (p=0.39). In Child B/C patients, the survival was considerably lower: 223 days (165-315 days) for cTACE versus 226 days (114-335 days) for DEB-TACE (p=0.53). CONCLUSION: The present study showed no significant difference in overall survival between cTACE and DEB-TACE in patients with HCC. However, the significantly lower number of treatments needed in the DEB-TACE group makes it a more appealing treatment option than cTACE for appropriately selected patients with unresectable HCC.

Efficacy and Safety of Vedolizumab in Patients with Inflammatory Bowel Disease in Association with Vedolizumab Drug Levels.

BACKGROUND: Vedolizumab (VDZ) is a well-established and important therapeutic option in the treatment of patients with inflammatory bowel disease (IBD). However, the significance of therapeutic drug monitoring (TDM) with VDZ remains a contradictory field in daily clinical practice. Our study aims to clarify the predictive impact of VDZ drug levels in long-term clinical outcomes in a real-world cohort. METHODS: Patients with moderate to severe ulcerative colitis (UC) and Crohn's disease (CD) from a tertiary IBD referral center at the University Hospital Augsburg, Germany, were enrolled in this single-center retrospective data analysis. Clinical and endoscopic data were collected at month 6, month 12, and at the last time of follow-up, and outcomes were correlated with VDZ levels at week 6. RESULTS: This study included 95 patients, 68.4% (n = 65) with UC, 24.2% (n = 23) with CD, and 7.4% (n = 7) with indeterminate colitis (CI). Patients with a mean VDZ treatment time of 17.83 months ± 14.56 showed clinical response in 29.5% (n = 28) and clinical remission in 45.3% (n = 43) at the end of the study. Endoscopic response occurred in 20.0% (n = 19) and endoscopic remission in 29.5% (n = 28) at the end of the study. The sustained beneficial effect of VDZ was also reflected in a significant change in biomarker levels. VDZ trough level at week 6 was determined in 48.4% (n = 46) with a mean of 41.79 µg/mL ± 24.58. A significant association between VDZ level at week 6 and both short and long-term outcomes could not be demonstrated. However, numerically higher VDZ levels were seen in patients with endoscopic and clinical improvement at month 6 and at the time of last follow-up. CONCLUSIONS: This study demonstrated efficacy and safety for VDZ in a real-world cohort. Although, for some parameters, a clear trend for higher VDZ levels at week 6 was seen, the efficacy of VDZ was not significantly correlated to VDZ level at week 6, which questions the predictive value of VDZ levels in the real world.

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.

Gastric cancer- endoscopic treatment of early lesions, the West learns from the East.

Gastric cancer still has one of the highest incidence rates worldwide. Screening programs have been established in high incidence regions, especially in Asia, but in the West, screening for gastric cancer is not generally recommended. Gastroscopy is the gold standard for diagnosing gastric cancer. For the treatment of early gastric cancer, endoscopic resection is the method of choice. With the ESD technique, larger lesions can be resected en-bloc. Guideline and extended guideline criteria for the choice of lesions for ESD have been evaluated extensively, initially in Asia and later in the West as well. For lesions which are out of indication, a surgical approach must be recommended. To detect early recurrence or metachronous lesions, follow-up should be performed after ER.

A technical review of artificial intelligence as applied to gastrointestinal endoscopy: clarifying the terminology.

Background and aim The growing number of publications on the application of artificial intelligence (AI) in medicine underlines the enormous importance and potential of this emerging field of research. In gastrointestinal endoscopy, AI has been applied to all segments of the gastrointestinal tract most importantly in the detection and characterization of colorectal polyps. However, AI research has been published also in the stomach and esophagus for both neoplastic and non-neoplastic disorders. The various technical as well as medical aspects of AI, however, remain confusing especially for non-expert physicians. This physician-engineer co-authored review explains the basic technical aspects of AI and provides a comprehensive overview of recent publications on AI in gastrointestinal endoscopy. Finally, a basic insight is offered into understanding publications on AI in gastrointestinal endoscopy.