Virgo Surgical Video Solutions has unveiled EndoML, an artificial intelligence development platform aimed at gastroenterologists and pharmaceutical researchers.

EndoML is designed to streamline the transition from AI concept to clinical application, enhancing the capabilities of professionals in the field.

The platform leverages Virgo’s AI foundation model, EndoDINO, which has been trained on a vast dataset of over 130,000 full-length gastrointestinal endoscopy videos.

The model has shown exceptional performance in various GI endoscopy applications.

Its applications include anatomical landmark recognition, polyp segmentation, and inflammatory bowel disease severity scoring.

Virgo CEO Matt Schwartz said: “The EndoML platform puts the power of artificial intelligence directly into the hands of clinicians and researchers, allowing them to harness the full power of AI in gastroenterology.

“Our goal is to streamline and democratize AI development in the field, facilitating quicker transitions from research innovations to tangible patient care improvements. I’ll be doing live demos of EndoML at DDW 2025.”

The EndoML platform allows researchers to train custom AI models using their unique datasets.

By leveraging EndoDINO as the backbone, researchers can achieve superior model performance with reduced need for labelled data.

Developed models are made accessible for ongoing research and collaboration, promoting further advancements within the scientific community.

Virgo has previously demonstrated EndoML’s accuracy and effectiveness in Mayo Endoscopic Scoring for both frame-level and video-level analyses.

Virgo engineering VP Chris Lawcock said: “EndoML not only simplifies AI model development but also provides pharmaceutical sponsors flexibility to train and deploy these models on their own terms.

“EndoML models seamlessly integrate within VirgoTrials, our data capture and management solution for clinical trials.

“At the same time, sponsors can also leverage EndoML models in their own data analysis environments. It’s a comprehensive toolset designed to meet the unique needs of clinical and pharmaceutical research.”