Table of Contents
Experts’ discussions during the ESMO Gastrointestinal Cancers Congress 2026 explored how AI is moving beyond image recognition to support surgery, pathology and image-guided intervention, reflecting a broader shift in AI GI cancer clinical value from technical performance toward measurable patient outcomes.
The Evidence Base Behind AI GI Cancer Clinical Value
The value of AI in gastrointestinal oncology is becoming increasingly clear, as demonstrated by recent advances. In colorectal cancer, the MSIntuit algorithm detected microsatellite instability directly from routine haematoxylin and eosin slides with 96-98% sensitivity, while deep learning models have been developed to predict lymph node metastasis from pre-treatment CT imaging in gastric cancer.
A New Phase Where Clinical Value Matters as Much as Performance
Experts speaking at the congress suggested the field is entering a new phase, where clinical value is proving to be just as important as technical performance.
Surgical AI Moves Beyond Anatomy Recognition
According to Dr. Marie Hanaoka, Institute of Science Tokyo, Japan, the field is moving towards “outcome-linked intelligence,” in which AI-derived measurements made during surgery are linked to clinically meaningful patient outcomes rather than simply recognising anatomy or surgical workflow.
A Concrete Example From Robotic Rectal Cancer Surgery
Presenting proof-of-concept data using the AI platform EUREKA X, Hanaoka quantified the preservation of loose connective tissue during robotic rectal cancer surgery. In a matched case-control study of 44 patients, AI-derived measures of tissue preservation were associated with postoperative urinary dysfunction. When combined with patient age, the model achieved an area under the curve of approximately 0.81. “Surgical AI is no longer just about recognition accuracy,” Hanaoka said. “We must ask whether this AI information is meaningful for our patients.” Looking ahead, she concluded: “The future is not AI versus surgeons. It is AI-enhanced surgeons.”
Seeing What CT Can’t: AI GI Cancer Clinical Value in Image-Guided Intervention
A similar evolution is taking place in image-guided intervention. Prof. Max Seidensticker, LMU Klinikum, Munich, Germany, described how MRI-guided interstitial brachytherapy is expanding treatment options for liver tumours that are unsuitable for thermal ablation. MRI guidance enables clinicians to visualise lesions that are often invisible on CT.
Results From the MR BRIGHT Study
In the prospective MR BRIGHT study, MRI-guided brachytherapy achieved significantly higher local recurrence-free survival than CT-guided treatment, 95.1% versus 79.9%, while substantially reducing low-dose radiation exposure to healthy liver tissue, 15% versus 43%. Looking ahead, Seidensticker described how machine learning algorithms could accelerate MRI image reconstruction and reduce artefacts, making real-time MRI guidance increasingly practical during intervention.
From Promising Algorithms to Clinical Tools in Pathology
Pathology may be where AI’s transition from research to routine clinical practice is most evident. Prof. Magali Svrcek, Sorbonne Université, Paris, France, described AI as the product of four converging developments: an expanding biomarker landscape, increasing pathology workloads, the widespread adoption of digital pathology and persistent diagnostic variability, together transforming the pathology slide into “a source of quantitative biological data.”
The ESMO Framework for AI-Based Biomarkers
Presenting the ESMO basic requirements for AI-based biomarkers in oncology, Svrcek explained how algorithms are progressing from standardising existing biomarkers to predicting molecular alterations directly from H&E slides and ultimately discovering entirely new biomarkers. She also highlighted emerging models capable of recognising when they should not make a prediction.
Why Demonstrating Clinical Value Remains the Central Challenge for AI GI Cancer Clinical Value
For Svrcek, “The real challenge is not to demonstrate the technical performance or robustness of the algorithm, but to demonstrate the clinical value.” Looking ahead, she concluded: “The future of pathology is not the replacement of pathologists by AI. It is the transformation of pathologists into experts in complex biological data, serving precision medicine and patient care.”
A Consistent Theme Across All Three Presentations
Across surgery, image-guided intervention, and pathology, the presentations at ESMO GI 2026 converge on a shared theme: AI’s next phase of development depends less on improving raw technical accuracy and more on proving that its outputs translate into meaningful, measurable improvements in patient outcomes.
What This AI GI Cancer Clinical Value Shift Means Going Forward
As GI oncology moves from validating AI’s technical recognition capabilities toward demonstrating outcome-linked clinical value, researchers and clinicians will likely face growing pressure to design studies, like Hanaoka’s matched case-control analysis and Seidensticker’s prospective MR BRIGHT trial, that directly connect AI outputs to patient-relevant endpoints rather than intermediate technical metrics alone. Given Svrcek’s framing of pathologists transforming into experts in complex biological data rather than being replaced, this shift may also reshape training and workforce expectations across oncology specialties adopting AI tools.
What to Watch Going Forward
As real-time MRI guidance and AI-enhanced surgical feedback systems continue maturing, industry observers will likely watch for additional prospective studies validating whether tools like EUREKA X and MRI-guided brachytherapy protocols deliver consistent outcome improvements across larger, more diverse patient populations beyond these initial proof-of-concept and single-institution studies. Given the shared emphasis across all three presentations on clinical value over technical performance, this AI GI cancer clinical value framework may increasingly shape how oncology AI tools are evaluated and adopted across surgical, interventional, and pathology applications in the years ahead.
For more healthcare industry updates, insights and news, visit DistilINFO. Click here to subscribe to stay informed.
