GI Genius Module 100 (GGM100.US); GI Genius Module 200 (GGM200.US); ColonPRO 4.0 (CPRO40.US)

The GI Genius™ system, with its updated ColonPRO™ 4.0 software, utilizes a retrained artificial intelligence neural network to provide real-time computer-assisted detection of colonic mucosal lesions during standard white-light endoscopy. This technology aims to improve the accuracy of polyp and adenoma identification by visually and audibly alerting endoscopists to suspicious areas, thereby enhancing diagnostic yield in colorectal cancer screening and surveillance.

The GI Genius™ system, comprising Modules 100/200 and ColonPRO™ 4.0 software, is an artificial intelligence-driven tool designed to assist endoscopists during standard white-light colonoscopies. Its primary purpose is to enhance the real-time detection of colonic mucosal lesions, such as polyps and adenomas, in patients undergoing screening or surveillance. The system operates by processing live colonoscopy video feeds. Utilizing a retrained neural network, the ColonPRO™ 4.0 software analyzes these images to identify regions consistent with colorectal lesions. When a potential lesion is detected, the system superimposes visual markers (green squares) and provides an audible alert on the endoscopic display monitor, drawing the clinician's attention to the area of interest. This computer-assisted detection is intended to supplement, not replace, the endoscopist's clinical judgment and is specifically for use with standard white-light endoscopy. The ColonPRO™ 4.0 software represents an update, most notably featuring an improved neural network achieved through retraining, which directly enhances its detection capabilities compared to its predecessor. Non-clinical testing demonstrated superior performance for ColonPRO™ 4.0, including an increased lesion-based sensitivity (88.07% from 86.5%), a higher true positive rate per frame (mean 60.14% from 58.30%), and a lower false positive rate per frame (mean 1.79% from 1.96%). Frame-level analysis further confirmed these improvements with more true positives and fewer false positives and false negatives. Additionally, the software update includes containerization for optimized deployment and a new non-clinical "Procedure Highlights" function for practice optimization, though this does not impact the core detection performance. The system maintains compatibility with standard endoscopy video processors and displays, including 4K UHD for Module 200.

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. How does the retraining of the neural network in ColonPRO™ 4.0 specifically contribute to the observed improvements in lesion-based sensitivity and the reduction in false positives/negatives compared to version 3.0.2?; What types of colonic mucosal lesions (e.g., size, morphology, histology) does the GI Genius™ system with ColonPRO™ 4.0 exhibit the highest and lowest detection efficacy for?; Beyond the reported metrics (sensitivity, false positive rate), what is the system's performance on differentiating subtle or flat lesions, which are often challenging to detect?; How does the 'Procedure Highlights' function collect and aggregate data, and what specific metrics does it provide to healthcare organizations for practice optimization, without compromising patient data privacy?; Considering the updated SOUP components (e.g., CUDA, TensorRT), what is the impact on the system's processing speed or latency in real-time lesion detection, and how does this compare to the previous version?