Introduction

Deep learning systems in gastrointestinal endoscopy are traditionally pretrained on large, diverse datasets comprising generic, non-medical images before being trained on scarce endoscopic data. Foundation models are models pretrained based on domain-specific datasets. For endoscopy, foundation models require large datasets of general endoscopic images. Yet datasets for developing such models remain limited. In this study, we present GastroNet-5M, a dataset comprising 4,820,653 endoscopic images of approximately 500,000 procedures.

Aims & Methods

GastroNet-5M is composed of anonymized endoscopic images collected from eight Dutch hospitals between 2012 and 2020. Using a self-supervised learning approach, this dataset enabled the development of a foundation model for various endoscopic AI applications. In this study, we compared GastroNet-5M-pretrained models against current state-of-the-art models with traditional pretraining approaches and all publicly available endoscopic foundation models.
We evaluated diagnostic performance on 9 classification and 9 segmentation tasks, including Barrett’s neoplasia detection, colorectal polyp characterization, and gastric cancer invasion depth prediction. All models were trained from scratch for each task. Data efficiency was assessed by repeating all classification tasks with progressively smaller training sets (i.e. 100%, 50%, 25%, 10% and 5%). Robustness was evaluated across 12 datasets reflecting data heterogeneity such as variation in endoscope manufacturer, imaging modality (e.g., virtual chromoendoscopy), and image quality.

Results

GastroNet-5M-pretrained models outperformed all benchmarks in 8 of 9 classification tasks, with an average AUC improvement of 6.8% (Table 1). For segmentation, they achieved higher Dice scores compared to all other models across all 9 tasks, with an average gain of 15.4%. In data efficiency experiments, GastroNet-5M models achieved the highest AUC in 5 of 9 tasks when training data was limited, while no other model led in more than one. In robustness testing, they outperformed other models in 8 of 12 tasks evaluating heterogeneous data conditions.

Table 1. Area-Under-the-Curve scores of the GastroNet-5M-pretrained model and benchmark models for all classification tasks.

Conclusion

GastroNet-5M, a multicenter dataset of over 5 million unlabeled endoscopic images, offers a valuable resource for pretraining deep learning models in endoscopy. The use of GastroNet-5M enhances diagnostic accuracy, reduces the need for scarce application-specific endoscopic imagery, and improves robustness of endoscopic AI systems to the inevitable data heterogeneity in clinical practice. GastroNet-5M is publicly accessible for further research and development.

References

1. Sanderson E, Matuszewski BJ. arXiv preprint arXiv:2401.06278. 2024.
2. van der Putten J, de Groof J, Struyvenberg M, et al. Artif Intell Med. 2020;107.
3. Wang Z, Liu C, Zhang S, et al. arXiv preprint arXiv:2306.16741. 2023.
4. Liu Z, Mao H, Wu C-Y, et al. arXiv preprint arXiv:2201.03545. 2022.
5. Dosovitskiy A, Beyer L, Kolesnikov A, et al. arXiv preprint arXiv:2010.11929. 2020.
6. Oquab M, Darcet T, Moutakanni T, et al. arXiv preprint arXiv:2304.07193. 2023.
7. Jocher G, Qiu J. Ultralytics YOLOv11, v11.0.0. 2024.
8. de Groof AJ, Struyvenberg MR, van der Putten J, et al. Gastroenterology. 2020;158(4):915–29.e4.
9. Fockens KN, Jukema JB, Boers T, et al. United European Gastroenterol J. 2023;11(4):324–36.
10. Jukema J, Kusters C, Jong M, et al. Endoscopy (ESGE Days). 2023.
11. GIANA Challenge. Available from: https://giana.grand-challenge.org
12. Borgli H, Thambawita V, Smedsrud PH, et al. Sci Data. 2020;7(1):283.
13. Polat G, Kani HT, Ergenc I, et al. Inflamm Bowel Dis. 2023;29(9):1431–9.
14. van der Zander QEW, Schreuder RM, Fonollà R, et al. Endoscopy. 2021.
15. Tang D, Zhou J, Wang L, et al. Front Oncol. 2021;11:654093.
16. Houwen BB, Hazewinkel Y, Giotis I, et al. Endoscopy. 2022;54(Suppl 1):eP133.

Introduction

The European Union has signalled gastric cancer screening and prevention as a major health priority. The TOGAS consortium (TOwards GAstric cancer Screening in the European Union; togas.lu.lv), funded in response to this call, investigates different strategies for primary and secondary prevention of gastric cancer in Europe. One pilot study aims at determining the yield of having an oesophagogastroduodenoscopy (OGD) on the same day of a colonoscopy for colorectal cancer screening or polyp surveillance.

Aims & Methods

This is the first interim data on n=602 individuals recruited in 8 centres across Europe (France, Germany, Ireland, Latvia, Lithuania, the Netherlands, Portugal, and Spain). Individuals of the age of 50-74 years, attending for screening or surveillance colonoscopy, were invited to participate if they did not have a past diagnosis of gastric preneoplastic conditions requiring surveillance. Standardization of endoscopic procedures was assured by a rigorous protocol including all quality parameters defined by ESGE. This included standards for image documentation and biopsy sampling. Demographics, comorbidities, medication and the previous surgical and family history were recorded. Patients with marked preneoplastic changes (i.e. glandular atrophy or intestinal metaplasia [IM] at OLGA/OLGIM stage 3 or 4, and/or EGGIM score >5) were registered as requiring further endoscopic surveillance. Mann-Whitney U-test was applied for comparison of age between groups and Chi2 test for comparison of categorical variables (significance for p<0.05).

Results

The mean age of participants was 62 years (SD +/- 8.0), with 55.8% being male. Of 245 individuals who underwent an OGD prior to inclusion, 93.1% of these procedures were done >3years before and 60.4% were done at a tertiary centre. We identified 6 (1.0%) malignant/neoplastic lesions, 4 gastric cancers/dysplastic lesions and 2 oesophageal carcinomas. A positive H. pylori status was diagnosed in 34.2%. Endoscopically, a diagnosis of preneoplastic conditions was made in 217 participants (36.0%). Of these, 84 (14.0%) showed gastric atrophy, 45 (7.5%) gastric IM and 75 (12.5%) both conditions combined. Of those with gastric atrophy, this was classified as multifocal in 44 (27.6%) individuals, of those with IM in 45 (37.5%). OLGA stages 3 and 4 were confirmed for 1.9% and 0.8%, OLGIM stages 3 and 4 for 2.3% and 0.4%. Benign ulcers were diagnosed in 28 (4.7%) participants. The majority being gastric ulcers in 19 (3.2%) individuals, duodenal ulcers in 9 (1.5%), and oesophageal ulcers in 2 (0.3%). An additional 13 (2%) participants were diagnosed with Barrett’s oesophagus. It is of note that also 12 (2.0%) colorectal cancers were identified in the cohort. A positive H. pylori status was associated with the need for gastric surveillance (p=0.031).

Conclusion

This interim analysis of the first European pilot study on a combined screening for upper and lower GI tumours shows that this approach is feasible. Regarding the detection of upper GI malignancy and preneoplastic conditions, our findings are in line with previous systematic reviews, but we also demonstrate a reasonable rate of non-malignant findings that are also of relevance when defining the best screening strategy in Europe.

Disclosure

JB: Advisory fee from Flynn Pharma Ltd, Juvise Pharmaceuticals.
MCWS: Research Support from Sysmex, Norgine, Sentinel
OK: Speakers fee, advisory fee from Abbvie, Pfizer, Takeda, Janssen, Shire.

Introduction

Immune checkpoint inhibitors (ICIs), a new class of immunotherapeutic agents, are the first-line treatment for many advanced epithelial cancers. Gastrointestinal irAEs, particularly diarrhea and colitis, affect approximately 40% of ICI-treated patients and often lead to therapy discontinuation.¹ High-dose corticosteroids represent the first-line treatment, but about one-third of patients show inadequate response and require second-line immunosuppressants such as Infliximab or Vedolizumab.² As the body of evidence about the role of gut microbiota in the pathogenesis of GI irAE grows, fecal microbiota transplantation (FMT) is emerging as a potential treatment for steroid- and biologic-refractory ICI-induced colitis.³ Although data are still preliminary, FMT represents a promising therapeutic strategy in this setting.

Aims & Methods

The aim of this study is to evaluate the safety and efficacy of FMT in patients referred to our Microbiome clinic for steroid- and biologic-refractory ICI-induced colitis. After clinical evaluation, patients were treated with FMT via colonoscopy from healthy donors. FMT was performed as already described, following international clinical guidelines.^4,5 Following the endoscopic procedure, patients were evaluated at 1, 4, 8, and 12 weeks. We collected data on: patient characteristics, cancer type, cancer treatment and its suspension, severity of diarrhea before FMT (graded via CTCAE)⁶, time in weeks from initiation of cancer therapy to onset of diarrhea, grade of diarrhea following treatment and resumption of ICI.

Results

Our cohort included 7 patients with steroid- and biologic-refractory ICI-induced colitis (grade 1: n=0 ; grade 2: n=1 ; grade 3: 5=x; grade 4: n=1). Four patients were affected by renal cell carcinoma, 2 by melanoma, 1 by pulmonary cancer. All patients were treated with PD-1/PDL-1 inhibitors or CTLA-4 ( Nivolumab: n= 3; Nivolumab and Ipilimumab: n=2; Durvalumab: n= 1; Pembrolizumab: n= 1) and 6 patients interrupted ICI due to gastrointestinal irAEs. All patients received single FMT from healthy donors via colonoscopy. Following FMT, diarrhea improved to grade ≤ 1 in 6 patients (85%) at week 1, 4, 8, and 12 after the procedure, suggesting a sustained effect of microbiome modulation in the medium term. One patient died during follow-up for reasons other than ICI-colitis.

Conclusion

Our preliminary findings suggest that FMT may be a promising therapeutic strategy for refractory ICI-colitis. Further, larger and randomized studies are warranted to confirm our data.

References

1. Dougan M, Wang Y, Rubio-Tapia A, Lim JK. AGA Clinical Practice Update on Diagnosis and Management of Immune Checkpoint Inhibitor Colitis and Hepatitis: Expert Review. Gastroenterology. 2021;160(4):1384-1393. doi:10.1053/j.gastro.2020.08.063
2. Haanen J, Obeid M, Spain L, et al. Management of toxicities from immunotherapy: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol. 2022;33(12):1217-1238. doi:10.1016/j.annonc.2022.10.001
3. Wang Y, Wiesnoski DH, Helmink BA, et al. Fecal microbiota transplantation for refractory immune checkpoint inhibitor-associated colitis. Nat Med. 2018;24(12):1804-1808. doi:10.1038/s41591-018-0238-9
4. Cammarota G, Ianiro G, Kelly CR, et al. International consensus conference on stool banking for faecal microbiota transplantation in clinical practice. Gut. 2019;68(12):2111-2121. doi:10.1136/gutjnl-2019-319548
5. Ianiro G, Rossi E, Thomas AM, et al. Faecal microbiota transplantation for the treatment of diarrhoea induced by tyrosine-kinase inhibitors in patients with metastatic renal cell carcinoma. Nat Commun. 2020;11(1):4333. doi:10.1038/s41467-020-18127-y
6. Freites-Martinez A, Santana N, Arias-Santiago S, Viera A. Using the Common Terminology Criteria for Adverse Events (CTCAE - Version 5.0) to Evaluate the Severity of Adverse Events of Anticancer Therapies. Actas Dermosifiliogr. 2021;112(1):90-92. doi:10.1016/j.ad.2019.05.009

Disclosure

G.I. has received personal fees for acting as speaker for Biocodex, Danone, Sofar, Malesci, Metagenics and Tillotts Pharma, and for acting as consultant and/or advisor for Ferring Therapeutics, Giuliani, Malesci and Tillotts Pharma. A.G. reports personal fees for consultancy from Eisai Srl, 3PSolutions, Real Time Meeting, Fondazione Istituto Danone, SinergieSrl, Board MRGE and Sanofi SpA personal fees for acting as a speaker for Takeda SpA, AbbVie and Sandoz SpA and personal fees for acting on advisory boards for VSL3 and Eisai. G.C. has received personal fees for acting as advisor for Ferring Therapeutics. All other authors have no conflicts of interest to disclose.

Introduction

Corpus atrophic gastritis (CAG) requires endoscopic-histological surveillance due to the risk of developing gastric neoplastic lesions (GNL). This study aimed to identify variables associated with GNL development at long-term follow-up using a feature ranking method based on a novel One-Class Support Vector Machine (SVM) model.

Aims & Methods

A dataset containing clinical, endoscopic, and histological variables from consecutive CAG patients (2001-2023) adhering to a surveillance program was considered. GNL presence at the longest available follow-up was recorded. Gastric biopsies and histological evaluations were conducted according to the updated Sydney system. CAG patients with any missing data or a GNL diagnosis at the baseline gastroscopy were excluded from the final analysis. A Fisher score-based feature ranking method and One-Class SVM were used to select the optimal subset of baseline variables associated with GNL development, aiming to achieve average sensitivity (Se) and specificity (Sp) above 80%.

Results

Overall, 355 CAG patients were initially considered. Of these, 36 were excluded due to the presence of GNL at baseline gastroscopy, and 216 for missing data. Thus, a total of 103 patients were considered and grouped into: CAG patients with [22 patients (F 68.1%), median age 68 (35-83) years] and without GNL at follow-up [81 patients (F 72.8%), median age 59(26-84) years]. After a median follow-up of 60 (12-192) months, 13 epithelial GNL (gastric adenocarcinoma or high/low-grade dysplasia) and nine type-1 gastric-neuroendocrine-tumors (T1gNET) were recorded. Parietal-cell-antibodies and pepsinogen-I < 30 μg/l were associated with both epithelial GNL and T1gNET. Antral inflammation and age > 60 were linked to epithelial GNL, while anti-thyroperoxidase-antibodies, smoking, and dyspeptic symptoms at CAG diagnosis were linked to T1gNET. Low-dose aspirin and H. pylori eradication therapy showed inverse associations with epithelial GNL and T1gNET, respectively.

Conclusion

This is the first study in which an AI model simultaneously considers clinical, endoscopic, and histological characteristics from a dataset of CAG patients, showing good sensitivity and specificity in identifying those variables associated with developing GNL. Nonetheless, we acknowledge the need for improved variable standardization and data balancing between groups to enhance model performance further.

Introduction

Capsule endoscopy (CE) is a minimally invasive and essential tool for assessing small bowel disorders, especially Crohn’s disease. Accurate detection of ulcers and erosions is vital for evaluating disease activity, guiding treatment, and monitoring therapeutic response. Although CE has revolutionized lesion detection, its manual interpretation remains time-consuming and subject to interobserver variability. Recent advances in artificial intelligence (AI), particularly convolutional neural networks (CNNs), offer promising solutions by enhancing diagnostic accuracy, reducing variability, and improving workflow efficiency. This study aimed to develop and validate a robust AI model capable of detecting and distinguishing small bowel ulcers and erosions across multiple CE platforms and clinical centers.

Aims & Methods

In a prospective, multicenter study conducted between January 2021 and April 2024, data were collected from centers in Portugal, Spain, and the United States. Two CE devices—PillCamSB3 and Olympus EC-10—were used to analyze 137 anonymized exams. The AI-assisted readings generated by a deep learning model were compared to standard-of-care (SoC) interpretations, with an expert consensus panel serving as the reference standard. Performance metrics included sensitivity, specificity, positive and negative predictive values (PPV, NPV), and area under the receiver operating characteristic curve (AUROC).

Results

Expert consensus identified ulcers and erosions in 56 patients (40.9%). Compared to the reference standard, SoC readings achieved 60.7% sensitivity, 98.8% specificity, 97.1% PPV, 78.4% NPV, and 83.2% accuracy. In contrast, the AI-assisted model achieved 94.6% sensitivity, 80.2% specificity, 76.8% PPV, 95.6% NPV, and 86.1% accuracy. The AI model was found to be both non-inferior (p < 0.001) and statistically superior (p < 0.001) to SoC. Notably, it identified 68 lesions, nearly double those detected by SoC (35), and demonstrated consistent performance across devices and centers. The mean AI-assisted reading time was 239 seconds (SD 138 seconds) per exam.

Conclusion

The AI-assisted model significantly outperformed conventional SoC in the detection of small bowel ulcers and erosions, achieving higher sensitivity and overall accuracy while substantially reducing reading time. This is the first multicenter validation of an interoperable AI solution for CE, capable of high diagnostic performance across different devices and clinical environments. These findings highlight the potential of AI to transform endoscopic evaluation and management in inflammatory bowel disease.