The EAGLE Trial, a multicenter randomized controlled study published in npj Digital Medicine, has demonstrated that Olympus's cloud-based CADDIE™ Computer-Aided Detection application significantly improves the detection of high-risk colorectal lesions during colonoscopy without compromising safety or workflow efficiency. This represents a pivotal development in colorectal cancer prevention, as the system specifically targets lesions most likely to progress to cancer but are traditionally difficult to detect.
Conducted across eight centers in four European countries with 841 patients and 22 endoscopists, the trial showed CADDIE-assisted colonoscopy achieved a 7.3% absolute increase in adenoma detection rate compared to standard colonoscopy. More importantly, the system demonstrated dramatic relative increases in detection of clinically significant lesion subtypes: 93% for large adenomas (>10 mm), 57% for non-polypoid adenomas, and 230% for sessile serrated lesions (SSLs). These findings address critical gaps in current colonoscopy practice, as flat lesions and SSLs are particularly challenging to identify yet carry substantial cancer risk.
The clinical importance of these results cannot be overstated. Sessile serrated lesions, which showed the most dramatic improvement in detection, are increasingly recognized as high-risk lesions whose identification is critical to reducing post-colonoscopy colorectal cancer. Recent research published in Lancet Gastroenterology & Hepatology has established the direct relationship between SSL detection and interval cancer risk reduction. The CADDIE application's training on a dataset specifically enriched with these hard-to-detect lesions explains its superior performance in this area.
Cloud deployment represents a fundamental shift in how AI tools reach clinical practice. By eliminating hardware dependencies through its cloud architecture, the system offers hospitals greater flexibility and enables subscription-based procurement models that could democratize access to advanced AI tools. This approach addresses concerns raised in recent guidelines from the European Society of Gastrointestinal Endoscopy about practical implementation barriers for AI-assisted colonoscopy.
Principal Investigator Rawen Kader noted that "cloud deployment can remove hardware barriers and give hospitals access to the latest AI innovations, which has the potential of improving detection of the lesions that matter most for reducing colorectal cancer risk." The complete study is available at https://doi.org/10.1038/s41746-025-02270-1.
Importantly, the trial demonstrated no increase in unnecessary resections, addressing safety concerns that have accompanied some AI implementations. This balance between improved detection and maintained safety standards aligns with quality indicators recently established in Gastrointestinal Endoscopy that emphasize both lesion detection and appropriate intervention. The system's real-time performance across diverse testing environments further supports its practical utility in routine clinical settings.
As colorectal cancer remains a leading cause of cancer mortality worldwide, with pathways of carcinogenesis well-documented in Gastroenterology research, tools that enhance early detection of precancerous lesions represent significant advances in preventive medicine. The CADDIE system's focus on clinically relevant lesions, combined with its accessible cloud-based deployment, positions it as a potentially transformative tool in the global effort to reduce colorectal cancer incidence and mortality through improved screening quality.
