International interobserver study confirms increase of diagnostic accuracy to 96% for mucinous vs. non-mucinous, and higher physician agreement with Cellvizio
AI + Cellvizio increases accuracy in detecting high-risk lesions from 53% to 87%
Paris and Boston, July 28, 2026 – 5:45 p.m. CEST – Mauna Kea Technologies (Euronext Growth: ALMKT), inventor of Cellvizio®, the multidisciplinary probe and needle-based confocal laser endomicroscopy (p/nCLE) platform, today announced that two new peer-reviewed studies published across leading gastroenterology journals reinforce the growing body of evidence supporting needle-based confocal laser endomicroscopy (nCLE) for the diagnosis of pancreatic cystic lesions (PCLs).
A new international interobserver study — Proven reproducibility across physicians
An international study, published in Gastrointestinal Endoscopy1, assessed how consistently physicians of differing experience interpret pancreatic cyst diagnostics. Adding Cellvizio imaging to standard evaluation increased mucinous-vs. non-mucinous accuracy to 96%, raised subtype-specific accuracy to approximately 85%, and significantly improved agreement among 15 independent observers. Reproducibility across clinicians is one of the keys to broad adoption.
A groundbreaking artificial intelligence (AI) study — Increasing accuracy in high-risk lesions
A clinical study, published in Techniques and Innovations in Gastrointestinal Endoscopy2, evaluated an nCLE-guided artificial intelligence algorithm for detecting high-grade dysplasia or early cancer within intraductal papillary mucinous neoplasms (IPMNs). In higher-risk, non-gastric IPMN subtypes, the AI algorithm achieved 87% accuracy versus 53% for conventional methods with comparable sensitivity across subtypes — early evidence supporting the Company’s roadmap to bring AI-assisted decision support to the Cellvizio platform.
Dr. Somashekar (Som) Krishna, Professor of Medicine and Director of Advanced Endoscopy at The Ohio State University Wexner Medical Center, commented: “In pancreatic cyst diagnosis, the cost of being wrong is enormous in either direction — we over-resect benign cysts and still miss cancers. In the ongoing multicenter CLIMB study evaluating EUS-nCLE, adding real-time EUS-nCLE to the standard evaluation raised diagnostic accuracy from 73% to 85%, reduced unnecessary surgery for benign cysts by more than 40%, and cut missed cancers roughly six-fold. What makes that compelling is who produced it: 15 international endosonographers new to the technique, with the gains holding across all of them. And we can now build on it, applying artificial intelligence to nCLE video detected high-grade dysplasia and cancer with 87% accuracy in the non-gastric IPMN subtypes that carry the highest risk of progression. Used within a multimodal, size-based approach, nCLE, which is now included in the ESGE guidelines, is becoming a key piece in solving the pancreatic cyst puzzle.”
Sacha Loiseau, Ph.D., Chairman and Chief Executive Officer of Mauna Kea Technologies, concluded: “This growing body of independent, peer-reviewed evidence confirms what physicians are seeing in practice: Cellvizio gives them a unique window into pancreatic cysts that no other technology offers. Every unnecessary surgery avoided and every cancer detected earlier give meaning to what we do. It also strengthens our conviction as we scale adoption across U.S. centers and bring AI-assisted diagnosis to the platform.”
1 The Impact of Confocal Endomicroscopy Combined with Standard Diagnostics in Evaluating Large Pancreatic Cystic Lesions: An International Interobserver Study, https://doi.org/10.1016/j.gie.2026.06.067
2 Preoperative Prediction of Subtype and Artificial Intelligence–Driven Endomicroscopy Detection of Advanced Neoplasia in Intraductal Papillary Mucinous Neoplasms. https://doi.org/10.1016/j.tige.2025.250956
