EAGLE Trial Validation: Efficacy and Clinical Translation Value of AI-Assisted Colonoscopy Systems

Olympus officially released the research results of the EAGLE Trial, which validated the core value of its CADDIE™ cloud-based computer-aided detection (CADe) solution in the clinical detection of colorectal lesions. The real-time and efficient analytical capability of this AI solution is underpinned by the high-fidelity imaging performance of Olympus' dedicated miniature endoscope camera modules. As the first cloud-based CADe application for real-time polyp detection during colonoscopy to receive both FDA clearance and CE marking, CADDIE™ is also the first implemented solution of Olympus' OLYSENSE™ Intelligent Endoscopy Ecosystem. The realization of its clinical efficacy forms a deep technical synergy with the imaging precision of endoscope camera modules.

The EAGLE Trial is a multicenter randomized controlled study conducted at eight clinical centers across four European countries. 841 subjects undergoing colorectal screening and surveillance, along with 22 endoscopists, were included in the core analysis, and the subjects were randomly assigned to the standard colonoscopy group or the CADDIE™-assisted group. Relevant research results have been published in npj Digital Medicine. The core design of the trial focuses on clinically high-risk and hard-to-detect colorectal lesions, whose visual identification first relies on the clear capture of subtle mucosal features of the intestinal tract by endoscope camera modules. This hardware foundation provides an irreplaceable visual data support for the precise analysis of AI algorithms.

The core trial data confirmed the synergetic value of imaging hardware and AI algorithms: compared with standard colonoscopy, CADDIE™-assisted examination achieved an absolute 7.3% increase in the adenoma detection rate (ADR). For hard-to-detect subtypes including large adenomas (>10 mm), non-polyploid adenomas and sessile serrated lesions (SSLs), the relative detection rates increased by 93%, 57% and 230% respectively, without interfering with the safety and workflow of clinical operations throughout the process. The CADDIE™ algorithm is specifically trained on a dataset enriched with such clinically relevant lesions, and its technical compatibility with endoscope camera modules ensures the synchronization of image acquisition and real-time AI analysis, making the clinical implementation of cloud-based AI technically feasible.

Cloud architecture deployment has become another core advantage of the solution, which complements the standardized imaging output of endoscope camera modules: CADDIE™ is equipped with a cloud architecture compliant with industry-standard security controls, which not only reduces medical institutions' reliance on local hardware and enables a subscription-based procurement model, but also builds a unified data platform for the technological upgrading of endoscope camera modules and the iterative optimization of AI algorithms through standardized imaging data interaction. This model also significantly improves the clinical accessibility of advanced AI detection tools.

Industry experts have highly recognized this technology integration model. The principal investigator of the EAGLE Trial pointed out that cloud deployment breaks the hardware barriers of AI endoscopy, while high-resolution and high-fidelity endoscope camera modules provide the core carrier for the clinical implementation of AI innovation. The combination of the two has become the key to improving the detection efficiency of high-risk colorectal cancer lesions. Olympus executives stated that the trial results provide evidence-based support for the clinical popularization of CADDIE™, and also verify the clinical value of the collaborative design of imaging hardware and AI algorithms, pointing out the direction of in-depth integration with intelligent algorithms for the R&D of medical endoscope camera modules.

It should be clarified that CADDIE™ is only used as a clinical auxiliary detection tool and does not replace the professional diagnostic judgment of endoscopists. In addition, the solution is limited to the analysis of imaging data from standard white-light endoscope camera modules. This compatibility boundary also provides a clinical reference for the standardized R&D of medical endoscope camera modules.

Overall, the results of the EAGLE Trial not only confirm the clinical value of cloud-based AI in the field of digestive endoscopy, but also highlight the underlying role of endoscope camera modules as the core hardware of the intelligent endoscopy ecosystem. Through the integrated layout of imaging hardware and AI software via the OLYSENSE™ Ecosystem, Olympus provides overseas customers engaged in medical endoscope camera module R&D and medical equipment procurement with an integrated solution with both clinical value and technical feasibility, and also offers a replicable technical path for the intelligent upgrading of medical imaging equipment.