Prospective, Multi-Institutional, Real-Time Next-Generation Sequencing of Pancreatic Cyst Fluid Reveals Diverse Genomic Alterations That Improve the Clinical Management of Pancreatic Cysts.

BACKGROUND & AIMS: Next-generation sequencing (NGS) of pancreatic cyst fluid is a useful adjunct in the assessment of patients with pancreatic cyst. However, previous studies have been retrospective or single institutional experiences. The aim of this study was to prospectively evaluate NGS on a multi-institutional cohort of patients with pancreatic cyst in real time. METHODS: The performance of a 22-gene NGS panel (PancreaSeq) was first retrospectively confirmed and then within a 2-year timeframe, PancreaSeq testing was prospectively used to evaluate endoscopic ultrasound-guided fine-needle aspiration pancreatic cyst fluid from 31 institutions. PancreaSeq results were correlated with endoscopic ultrasound findings, ancillary studies, current pancreatic cyst guidelines, follow-up, and expanded testing (Oncomine) of postoperative specimens. RESULTS: Among 1933 PCs prospectively tested, 1887 (98%) specimens from 1832 patients were satisfactory for PancreaSeq testing. Follow-up was available for 1216 (66%) patients (median, 23 months). Based on 251 (21%) patients with surgical pathology, mitogen-activated protein kinase/GNAS mutations had 90% sensitivity and 100% specificity for a mucinous cyst (positive predictive value [PPV], 100%; negative predictive value [NPV], 77%). On exclusion of low-level variants, the combination of mitogen-activated protein kinase/GNAS and TP53/SMAD4/CTNNB1/mammalian target of rapamycin alterations had 88% sensitivity and 98% specificity for advanced neoplasia (PPV, 97%; NPV, 93%). Inclusion of cytopathologic evaluation to PancreaSeq testing improved the sensitivity to 93% and maintained a high specificity of 95% (PPV, 92%; NPV, 95%). In comparison, other modalities and current pancreatic cyst guidelines, such as the American Gastroenterology Association and International Association of Pancreatology/Fukuoka guidelines, show inferior diagnostic performance. The sensitivities and specificities of VHL and MEN1/loss of heterozygosity alterations were 71% and 100% for serous cystadenomas (PPV, 100%; NPV, 98%), and 68% and 98% for pancreatic neuroendocrine tumors (PPV, 85%; NPV, 95%), respectively. On follow-up, serous cystadenomas with TP53/TERT mutations exhibited interval growth, whereas pancreatic neuroendocrine tumors with loss of heterozygosity of ≥3 genes tended to have distant metastasis. None of the 965 patients who did not undergo surgery developed malignancy. Postoperative Oncomine testing identified mucinous cysts with BRAF fusions and ERBB2 amplification, and advanced neoplasia with CDKN2A alterations. CONCLUSIONS: PancreaSeq was not only sensitive and specific for various pancreatic cyst types and advanced neoplasia arising from mucinous cysts, but also reveals the diversity of genomic alterations seen in pancreatic cysts and their clinical significance.

Consensus statements on the current landscape of artificial intelligence applications in endoscopy, addressing roadblocks, and advancing artificial intelligence in gastroenterology.

BACKGROUND AND AIMS: The American Society for Gastrointestinal Endoscopy (ASGE) AI Task Force along with experts in endoscopy, technology space, regulatory authorities, and other medical subspecialties initiated a consensus process that analyzed the current literature, highlighted potential areas, and outlined the necessary research in artificial intelligence (AI) to allow a clearer understanding of AI as it pertains to endoscopy currently. METHODS: A modified Delphi process was used to develop these consensus statements. RESULTS: Statement 1: Current advances in AI allow for the development of AI-based algorithms that can be applied to endoscopy to augment endoscopist performance in detection and characterization of endoscopic lesions. Statement 2: Computer vision-based algorithms provide opportunities to redefine quality metrics in endoscopy using AI, which can be standardized and can reduce subjectivity in reporting quality metrics. Natural language processing-based algorithms can help with the data abstraction needed for reporting current quality metrics in GI endoscopy effortlessly. Statement 3: AI technologies can support smart endoscopy suites, which may help optimize workflows in the endoscopy suite, including automated documentation. Statement 4: Using AI and machine learning helps in predictive modeling, diagnosis, and prognostication. High-quality data with multidimensionality are needed for risk prediction, prognostication of specific clinical conditions, and their outcomes when using machine learning methods. Statement 5: Big data and cloud-based tools can help advance clinical research in gastroenterology. Multimodal data are key to understanding the maximal extent of the disease state and unlocking treatment options. Statement 6: Understanding how to evaluate AI algorithms in the gastroenterology literature and clinical trials is important for gastroenterologists, trainees, and researchers, and hence education efforts by GI societies are needed. Statement 7: Several challenges regarding integrating AI solutions into the clinical practice of endoscopy exist, including understanding the role of human-AI interaction. Transparency, interpretability, and explainability of AI algorithms play a key role in their clinical adoption in GI endoscopy. Developing appropriate AI governance, data procurement, and tools needed for the AI lifecycle are critical for the successful implementation of AI into clinical practice. Statement 8: For payment of AI in endoscopy, a thorough evaluation of the potential value proposition for AI systems may help guide purchasing decisions in endoscopy. Reliable cost-effectiveness studies to guide reimbursement are needed. Statement 9: Relevant clinical outcomes and performance metrics for AI in gastroenterology are currently not well defined. To improve the quality and interpretability of research in the field, steps need to be taken to define these evidence standards. Statement 10: A balanced view of AI technologies and active collaboration between the medical technology industry, computer scientists, gastroenterologists, and researchers are critical for the meaningful advancement of AI in gastroenterology. CONCLUSIONS: The consensus process led by the ASGE AI Task Force and experts from various disciplines has shed light on the potential of AI in endoscopy and gastroenterology. AI-based algorithms have shown promise in augmenting endoscopist performance, redefining quality metrics, optimizing workflows, and aiding in predictive modeling and diagnosis. However, challenges remain in evaluating AI algorithms, ensuring transparency and interpretability, addressing governance and data procurement, determining payment models, defining relevant clinical outcomes, and fostering collaboration between stakeholders. Addressing these challenges while maintaining a balanced perspective is crucial for the meaningful advancement of AI in gastroenterology.

A Combined DNA/RNA-based Next-Generation Sequencing Platform to Improve the Classification of Pancreatic Cysts and Early Detection of Pancreatic Cancer Arising From Pancreatic Cysts.

OBJECTIVE: We report the development and validation of a combined DNA/RNA next-generation sequencing (NGS) platform to improve the evaluation of pancreatic cysts. BACKGROUND AND AIMS: Despite a multidisciplinary approach, pancreatic cyst classification, such as a cystic precursor neoplasm, and the detection of high-grade dysplasia and early adenocarcinoma (advanced neoplasia) can be challenging. NGS of preoperative pancreatic cyst fluid improves the clinical evaluation of pancreatic cysts, but the recent identification of novel genomic alterations necessitates the creation of a comprehensive panel and the development of a genomic classifier to integrate the complex molecular results. METHODS: An updated and unique 74-gene DNA/RNA-targeted NGS panel (PancreaSeq Genomic Classifier) was created to evaluate 5 classes of genomic alterations to include gene mutations (e.g., KRAS, GNAS, etc.), gene fusions and gene expression. Further, CEA mRNA ( CEACAM5 ) was integrated into the assay using RT-qPCR. Separate multi-institutional cohorts for training (n=108) and validation (n=77) were tested, and diagnostic performance was compared to clinical, imaging, cytopathologic, and guideline data. RESULTS: Upon creation of a genomic classifier system, PancreaSeq GC yielded a 95% sensitivity and 100% specificity for a cystic precursor neoplasm, and the sensitivity and specificity for advanced neoplasia were 82% and 100%, respectively. Associated symptoms, cyst size, duct dilatation, a mural nodule, increasing cyst size, and malignant cytopathology had lower sensitivities (41-59%) and lower specificities (56-96%) for advanced neoplasia. This test also increased the sensitivity of current pancreatic cyst guidelines (IAP/Fukuoka and AGA) by >10% and maintained their inherent specificity. CONCLUSIONS: PancreaSeq GC was not only accurate in predicting pancreatic cyst type and advanced neoplasia but also improved the sensitivity of current pancreatic cyst guidelines.

Endoscopic management of gastrointestinal wall defects, fistula closure, and stent fixation using through-the-scope tack and suture system.

BACKGROUND : Multiple devices are available for tissue approximation. A new through-the-scope suturing (TTSS) device has recently been introduced; however, data on its scope of use and clinical effectiveness are limited. We aimed to assess the clinical course and effectiveness of this TTSS device. METHODS : A retrospective review was performed for consecutive patients who underwent TTSS application. Primary outcomes were technical and clinical success, and secondary outcomes included adverse events and long-term clinical success. RESULTS : 53 patients (mean age 67.8 years; 69.8 % females) were included, with a mean defect size of 32.6 mm (SD 11.9). Technical success was achieved in 51 patients (96.2 %). Clinical success was achieved in 49 patients (92.4 %). Two patients (3.8 %) experienced failed fistula closure after technical success. Long-term follow-up (> 30 days) was available for 45 patients (84.9 %), with a mean follow-up of 7.2 months. One patient (1.9 %) had self-reported bleeding that did not require further intervention. CONCLUSIONS : TTTS was an effective and safe method for the closure of large gastrointestinal defects and could be used for fistula closure and stent fixation, making it a valuable addition to the armamentarium of endoscopic closure devices.

The Modified Pancreatitis Activity Scoring System Shows Distinct Trajectories in Acute Pancreatitis: An International Study.

BACKGROUND & AIMS: The aims of this study were to: (1) assess the performance of the Pancreatitis Activity Scoring System (PASS) in a large intercontinental cohort of patients with acute pancreatitis (AP); and (2) investigate whether a modified PASS (mPASS) yields a similar predictive accuracy and produces distinct early trajectories between severity subgroups. METHODS: Data was prospectively collected through the Acute Pancreatitis Patient Registry to Examine Novel Therapies In Clinical Experience (APPRENTICE) consortium (2015-2018) involving 22 centers from 4 continents. AP severity was categorized per the revised Atlanta classification. PASS trajectories were compared between the three severity groups using the generalized estimating equations model. Four mPASS models were generated by modifying the morphine equivalent dose (MED), and their trajectories were compared. RESULTS: A total of 1393 subjects were enrolled (median age, 49 years; 51% males). The study cohort included 950 mild (68.2%), 315 (22.6%) moderately severe, and 128 (9.2%) severe AP. Mild cases had the lowest PASS at each study time point (all P < .001). A subset of patients with outlier admission PASS values was identified. In the outlier group, 70% of the PASS variation was attributed to the MED, and 66% of these patients were from the United States centers. Among the 4 modified models, the mPASS-1 (excluding MED from PASS) demonstrated high performance in predicting severe AP with an area under the receiver operating characteristic curve of 0.88 (vs area under the receiver operating characteristic of 0.83 in conventional PASS) and produced distinct trajectories with distinct slopes between severity subgroups (all P < .001). CONCLUSION: We propose a modified model by removing the MED component, which is easier to calculate, predicts accurately severe AP, and maintains significantly distinct early trajectories.

The relationship between pre-existing diabetes mellitus and the severity of acute pancreatitis: Report from a large international registry.

BACKGROUND/OBJECTIVES: The relationship between pre-existing diabetes mellitus (DM) and acute pancreatitis (AP) severity has not been established. We assessed the impact of pre-existing DM on AP severity in an international, prospectively ascertained registry. METHODS: APPRENTICE registry prospectively enrolled 1543 AP patients from 22 centers across 4 continents (8 US, 6 Europe, 5 Latin America, 3 India) between 2015 and 2018, and collected detailed clinical information. Pre-existing DM was defined a diagnosis of DM prior to AP admission. The primary outcome was AP severity defined by the Revised Atlanta Classification (RAC). Secondary outcomes were development of systemic inflammatory response syndrome (SIRS) or intensive care unit (ICU) admission. RESULTS: Pre-existing DM was present in 270 (17.5%) AP patients, of whom 252 (93.3%) had type 2 DM. Patients with pre-existing DM were significantly (p < 0.05) older (55.8 ± 16 vs. 48.3 ± 18.7 years), more likely to be overweight (BMI 29.5 ± 7 vs. 27.2 ± 6.2), have hypertriglyceridemia as the etiology (15% vs. 2%) and prior AP (33 vs. 24%). Mild, moderate, and severe AP were noted in 66%, 23%, and 11% of patients, respectively. On multivariable analysis, pre-existing DM did not significantly impact AP severity assessed by the RAC (moderate-severe vs. mild AP, OR = 0.86, 95% CI 0.63-1.18; severe vs. mild-moderate AP, OR = 1.05, 95% CI, 0.67-1.63), development of SIRS, or the need for ICU admission. No interaction was noted between DM status and continent. CONCLUSION: About one in 5 patients with AP have pre-existing DM. Once confounding risk factors are considered, pre-existing DM per se is not a risk factor for severe AP.

Endoscopic full-thickness resection of polyps involving the appendiceal orifice: a multicenter international experience.

BACKGROUND: Endoscopic resection of lesions involving the appendiceal orifice remains a challenge. We aimed to report outcomes with the full-thickness resection device (FTRD) for the resection of appendiceal lesions and identify factors associated with the occurrence of appendicitis. METHODS: This was a retrospective study at 18 tertiary-care centers (USA 12, Canada 1, Europe 5) between November 2016 and August 2020. Consecutive patients who underwent resection of an appendiceal orifice lesion using the FTRD were included. The primary outcome was the rate of R0 resection in neoplastic lesions, defined as negative lateral and deep margins on post-resection histologic evaluation. Secondary outcomes included the rates of: technical success (en bloc resection), clinical success (technical success without need for further surgical intervention), post-resection appendicitis, and polyp recurrence. RESULTS: 66 patients (32 women; mean age 64) underwent resection of colonic lesions involving the appendiceal orifice (mean [standard deviation] size, 14.5 (6.2) mm), with 40 (61 %) being deep, extending into the appendiceal lumen. Technical success was achieved in 59/66 patients (89 %), of which, 56 were found to be neoplastic lesions on post-resection pathology. Clinical success was achieved in 53/66 (80 %). R0 resection was achieved in 52/56 (93 %). Of the 58 patients in whom EFTR was completed who had no prior history of appendectomy, appendicitis was reported in 10 (17 %), with six (60 %) requiring surgical appendectomy. Follow-up colonoscopy was completed in 41 patients, with evidence of recurrence in five (12 %). CONCLUSIONS: The FTRD is a promising non-surgical alternative for resecting appendiceal lesions, but appendicitis occurs in 1/6 cases.

Association Between Portal Vein Thrombosis and Pouchitis in Patients with Ulcerative Colitis.

BACKGROUND: Pouchitis is the most common long-term complication in patients requiring colectomy ileal pouch-anal anastomosis with medically refractory ulcerative colitis or colitis-associated neoplasia. A previous small case series suggests associated between portal vein thrombosis (PVT) and ischemic pouchitis. AIM: To evaluate the association between PVT and other demographic and clinical factors and pouchitis. METHODS: We used Explorys Inc., a population-based database, to search medical records between 1999 and 2020 with SNOMED-CT code criteria for "construction of pouch" and "ileal pouchitis." Patients with pouchitis were compared to those with previous pouch construction without pouchitis. Factors associated with pouchitis identified with univariable analysis were introduced into a multivariable model. A post hoc analysis further stratified demographical findings of the association between PVT and pouchitis. RESULTS: We identified 7900 patients with ileal pouchitis (7.5%) and 97,510 with pouch construction without pouchitis. In multivariate binary logistic regression, adjusted odds ratio (aOR) for the risk of pouchitis in patients with PVT was 10.78 (95% confidence interval [CI] 7.04-16.49, P < 0.001). Other significant factors associated with pouchitis included male gender (aOR 1.11, 95% CI 1.02-1.21, P = 0.018), deep vein thrombosis (aOR 1.46, 95% CI 1.23-1.72, P < 0.001), and the use of non-steroidal anti-inflammatory drugs (aOR 1.37, 95% CI 1.28-1.45, P < 0.001). Smoking was a protective factor (aOR 0.30, 95% CI 0.33-0.36, P < 0.001). Further sub-analysis showed a higher prevalence of younger patients with PVT and pouchitis. CONCLUSIONS: We report PVT as an independent risk factor associated with pouchitis. Our findings support that PVT is a potentially manageable perioperative complication, and intervention may reduce the risk of pouchitis.

Primary obesity surgery endoluminal (POSE) for the treatment of obesity: a systematic review and meta-analysis.

BACKGROUND: Primary obesity surgery endoluminal (POSE) utilizes an incision-less operating platform system to create full-thickness plications in the gastric fundus and body (original POSE). Many studies have demonstrated the safety and efficacy of original POSE for the treatment of obesity. OBJECTIVE: We aimed to conduct a systematic review and meta-analysis of available literature in an attempt to evaluate the outcomes of original POSE per the ASGE task force thresholds. METHODS: Bibliographic databases were systematically searched for studies assessing the outcomes of POSE for the treatment of obesity. All randomized controlled trials (RCTs) and observational studies that assessed outcomes of POSE were included. Studies were included if they reported percent total weight loss (%TWL) or percent excess weight loss (%EWL) and the incidence of serious adverse events (SAE). RESULTS: A total of seven studies with 613 patients were included. Two included studies were RCTs, while the remaining were observational studies. Pooled mean %EWL at 3-6 months and 12-15 months were 42.62 (95% CI 37.56-47.68) and 48.86 (95% CI 42.31-55.41), respectively. Pooled mean %TWL at 3-6 months and 12-15 months was 13.45 (95% CI 8.93-17.97) and 12.68 (95% CI 8.13-17.23), respectively. Subgroup analysis of two RCTs showed that weight loss at 1 year was significantly higher in POSE patients (%EWL difference in means 19.45 (95% CI 4.65-34.24, p value = 0.01). The overall incidence of serious adverse events was only 2.84% and included GI bleeding, extra-gastric bleeding, hepatic abscess, severe pain, severe nausea, and severe vomiting. The mean number of total anchors placed in the fundus and body was 13.18 (95% CI 11.77-14.58), and the mean procedure time was 44.55 min (95% CI 36.44-52.65). CONCLUSION: POSE, a minimally invasive endoscopic bariatric therapy, is a safe and effective modality for the treatment of obesity. The outcomes of POSE meet and surpass the ASGE joint task force thresholds. Future studies should evaluate newer versions of this procedure that emphasize gastric body plication sparing the fundus.

Dynamic changes in the pancreatitis activity scoring system during hospital course in a multicenter, prospective cohort.

BACKGROUND AND AIM: The primary aim was to validate the Pancreatitis Activity Scoring System (PASS) in a multicenter prospectively ascertained acute pancreatitis (AP) cohort. Second, we investigated the association of early PASS trajectories with disease severity and length of hospital stay (LOS). METHODS: Data were prospectively collected through the APPRENTICE consortium (2015-2018). AP severity was categorized based on revised Atlanta classification. Delta PASS (ΔPASS) was calculated by subtracting activity score from baseline value. PASS trajectories were compared between severity subsets. Subsequently, the cohort was subdivided into three LOS subgroups as short (S-LOS): 2-3 days; intermediate (I-LOS): 3-7 days; and long (L-LOS): ≥7 days. The generalized estimating equations model was implemented to compare PASS trajectories. RESULTS: There were 434 subjects analyzed including 322 (74%) mild, 86 (20%) moderately severe, and 26 (6%) severe AP. Severe AP subjects had the highest activity levels and the slowest rate of decline in activity (P = 0.039). Focusing on mild AP, L-LOS subjects (34%) had 28 points per day slower decline; whereas, S-LOS group (13%) showed 34 points per day sharper decrease compared with I-LOS (53%; P < 0.001). We noticed an outlier subset with a median admission-PASS of 466 compared with 140 in the rest. Morphine equivalent dose constituted 80% of the total PASS in the outliers (median morphine equivalent dose score = 392), compared with only 25% in normal-range subjects (score = 33, P value < 0.001). CONCLUSIONS: This study highlighted that PASS can quantify AP activity. Significant differences in PASS trajectories were found both in revised Atlanta classification severity and LOS groups, which can be harnessed in AP monitoring/management (ClincialTrials.gov number, NCT03075618).

Prospective evaluation of an assessment tool for technical performance of duodenoscopes.

OBJECTIVE: While single-use and detachable-tip duodenoscopes have been recently developed to overcome risks of infection transmission, there are no reliable tools to objectively assess their technical performance. We evaluated the reliability and validity of a newly developed tool to assess the technical performance of reusable duodenoscopes. METHODS: An assessment tool was developed to measure duodenoscope performance based on three distinct criteria: maneuverability, mechanical/imaging characteristics and ability to perform requisite interventions. The assessment tool was tested prospectively on duodenoscopes used in endoscopic retrograde cholangiopancreatography (ERCP) procedures at nine academic medical centers over a 6-month period. The main outcome was reliability of the duodenoscope assessment tool, which was estimated using Cronbach's coefficient alpha (α). The secondary outcome was validity of the assessment tool. RESULTS: The assessment tool evaluated technical performance of reusable duodenoscopes in 1080 ERCP procedures. Indications were biliary in 92.8% and pancreatic in 7.2% procedures. The overall Cronbach's coefficient α for maneuverability was 0.81, assessment of mechanical/imaging characteristics was 0.92, and ability to perform requisite interventions was 0.87. On multiple linear regression analysis, prolonged procedure duration, older patient age and pancreatic interventions were significantly positively associated with higher (worse) scores. CONCLUSIONS: The newly developed assessment tool appears reliable and valid for evaluating the technical performance of duodenoscopes. Registration: ClinicalTrials.gov Identifier: NCT04004533.

Worldwide Variations in Demographics, Management, and Outcomes of Acute Pancreatitis.

BACKGROUND & AIMS: Few studies have compared regional differences in acute pancreatitis. We analyzed data from an international registry of patients with acute pancreatitis to evaluate geographic variations in patient characteristics, management, and outcomes. METHODS: We collected data from the APPRENTICE registry of patients with acute pancreatitis, which obtains information from patients in Europe (6 centers), India (3 centers), Latin America (5 centers), and North America (8 centers) using standardized questionnaires. Our final analysis included 1612 patients with acute pancreatitis (median age, 49 years; 53% male, 62% white) enrolled from August 2015 through January 2018. RESULTS: Biliary (45%) and alcoholic acute pancreatitis (21%) were the most common etiologies. Based on the revised Atlanta classification, 65% of patients developed mild disease, 23% moderate, and 12% severe. The mean age of patients in Europe (58 years) was older than mean age for all 4 regions (46 years) and a higher proportion of patients in Europe had comorbid conditions (73% vs 50% overall). The predominant etiology of acute pancreatitis in Latin America was biliary (78%), whereas alcohol-associated pancreatitis accounted for the highest proportion of acute pancreatitis cases in India (45%). Pain was managed with opioid analgesics in 93% of patients in North America versus 27% of patients in the other 3 regions. Cholecystectomies were performed at the time of hospital admission for most patients in Latin America (60% vs 15% overall). A higher proportion of European patients with severe acute pancreatitis died during the original hospital stay (44%) compared with the other 3 regions (15%). CONCLUSIONS: We found significant variation in demographics, etiologies, management practices, and outcomes of acute pancreatitis worldwide. ClinicalTrials.gov number: NCT03075618.

Clinical features of hypertriglyceridemia-induced acute pancreatitis in an international, multicenter, prospective cohort (APPRENTICE consortium).

BACKGROUND: The clinical features and outcomes of hypertriglyceridemia-induced acute pancreatitis (HTG-AP) are not well-established. OBJECTIVE: To evaluate the clinical characteristics of HTG-AP in an international, multicenter prospective cohort. METHODS: Data collection was conducted prospectively through APPRENTICE between 2015 and 2018. HTG-AP was defined as serum TG levels >500 mg/dl in the absence of other common etiologies of AP. Three multivariate logistic regression models were performed to assess whether HTG-AP is associated with SIRS positive status, ICU admission and/or moderately-severe/severe AP. RESULTS: 1,478 patients were included in the study; 69 subjects (4.7%) were diagnosed with HTG-AP. HTG-AP patients were more likely to be younger (mean 40 vs 50 years; p < 0.001), male (67% vs 52%; p = 0.018), and with a higher BMI (mean 30.4 vs 27.5 kg/m2; p = 0.0002). HTG-AP subjects reported more frequent active alcohol use (71% vs 49%; p < 0.001), and diabetes mellitus (59% vs 15%; p < 0.001). None of the above risk factors/variables was found to be independently associated with SIRS positive status, ICU admission, or severity in the multivariate logistic regression models. These results were similar when including only the 785 subjects with TG levels measured within 48 h from admission. CONCLUSION: HTG-AP was found to be the 4th most common etiology of AP. HTG-AP patients had distinct baseline characteristics, but their clinical outcomes were similar compared to other etiologies of AP.

Proceedings from the First Global Artificial Intelligence in Gastroenterology and Endoscopy Summit.

BACKGROUND AND AIMS: Artificial intelligence (AI), specifically deep learning, offers the potential to enhance the field of GI endoscopy in areas ranging from lesion detection and classification to quality metrics and documentation. Progress in this field will be measured by whether AI implementation can lead to improved patient outcomes and more efficient clinical workflow for GI endoscopists. The aims of this article are to report the findings of a multidisciplinary group of experts focusing on issues in AI research and applications related to gastroenterology and endoscopy, to review the current status of the field, and to produce recommendations for investigators developing and studying new AI technologies for gastroenterology. METHODS: A multidisciplinary meeting was held on September 28, 2019, bringing together academic, industry, and regulatory experts in diverse fields including gastroenterology, computer and imaging sciences, machine learning, computer vision, U.S. Food and Drug Administration, and the National Institutes of Health. Recent and ongoing studies in gastroenterology and current technology in AI were presented and discussed, key gaps in knowledge were identified, and recommendations were made for research that would have the highest impact in making advances and implementation in the field of AI to gastroenterology. RESULTS: There was a consensus that AI will transform the field of gastroenterology, particularly endoscopy and image interpretation. Powered by advanced machine learning algorithms, the use of computer vision in endoscopy has the potential to result in better prediction and treatment outcomes for patients with gastroenterology disorders and cancer. Large libraries of endoscopic images, "EndoNet," will be important to facilitate development and application of AI systems. The regulatory environment for implementation of AI systems is evolving, but common outcomes such as colon polyp detection have been highlighted as potential clinical trial endpoints. Other threshold outcomes will be important, as well as clarity on iterative improvement of clinical systems. CONCLUSIONS: Gastroenterology is a prime candidate for early adoption of AI. AI is rapidly moving from an experimental phase to a clinical implementation phase in gastroenterology. It is anticipated that the implementation of AI in gastroenterology over the next decade will have a significant and positive impact on patient care and clinical workflows. Ongoing collaboration among gastroenterologists, industry experts, and regulatory agencies will be important to ensure that progress is rapid and clinically meaningful. However, several constraints and areas will benefit from further exploration, including potential clinical applications, implementation, structure and governance, role of gastroenterologists, and potential impact of AI in gastroenterology.