GI Genius System 100 and GI Genius System 200

The GI Genius System is an artificial intelligence-driven tool that aids endoscopists by processing colonoscopy images in real-time to highlight potential colonic mucosal lesions using visual and auditory cues. The updated system introduces 4K UHD video compatibility through a specialized internal software library and an optional footswitch for enhanced usability, while maintaining its core AI-based lesion detection capabilities and performance characteristics.

The GI Genius System is a computer-assisted reading tool designed to support endoscopists by enhancing the detection of colonic mucosal lesions, such as polyps and adenomas, in real-time during standard white-light endoscopy examinations. The system utilizes artificial intelligence to process live colonoscopy video feeds, and upon identifying a potential lesion, it superimposes green square markers on the endoscopic display monitor, accompanied by a short audible alert. This functionality is intended for screening and surveillance endoscopic mucosal evaluations and is explicitly limited to use with standard white-light endoscopy imaging; it is not designed to replace the clinical decision-making process of the healthcare professional. This submission details design changes to the GI Genius System, resulting in software version 3.0.2 and the introduction of two distinct models: the GI Genius System 100 (an updated version of the current model previously cleared under K211951) and the GI Genius System 200 (a new model with 4K UHD capability). The System 200 incorporates a new videoboard compatible with 4K UHD video processors like the Olympus CV-1500 EVIS X1. To maintain the integrity of the existing AI detection software, a new internal "wrapper" library is implemented; this library converts the 4K UHD video input to HD for the AI processing and then upscales the AI's HD screen overlay to 4K for display. Both systems now offer compatibility with optional accessories, including a footswitch and USB K-switch, allowing users to toggle the lesion markers. Comprehensive non-clinical testing confirmed the system's safety and performance, including non-inferiority of the 4K system compared to HD, and verified electrical safety and electromagnetic compatibility for the new hardware and accessories.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What is the precise impact of the "wrapper" library's HD-to-4K upscaling on the perceived sharpness or accuracy of the lesion marker overlay when using a 4K UHD video source with the GI Genius System 200?; How does the real-world clinical performance of the GI Genius System 200 with 4K UHD video compare to the System 100 (or predicate HD system) in terms of polyp detection rate and adenoma detection rate, beyond the per-frame non-inferiority testing on pre-recorded videos?; What specific user feedback or clinical needs prompted the development of the footswitch functionality for toggling lesion markers, and how has its usability been evaluated in a simulated or actual clinical workflow?; Given the system is trained on images with lesions like polyps, including flat morphology, what is the range of lesion sizes and specific morphologies for which the AI demonstrates optimal sensitivity and specificity according to the unchanged performance data from K211951?; How does the documented video signal processing delay of 0.74 µs for the GI Genius System 200, an improvement over the predicate, practically influence the real-time user experience for the endoscopist during a procedure?