AI’s Transformative Role in Colorectal Cancer Screening

In the dynamic field of gastroenterology, AI is revolutionizing clinical practices, particularly in colorectal cancer screening. Dror Zur, CEO of Magentiq Eye shares insights on how AI enhances adenoma detection rates (ADR) and its broader implications. This discussion explores AI’s integration into clinical workflows, its impact on physician practices, and the future of AI-driven medical procedures.

Moe: Adenoma detection is key for colorectal cancer outcomes. How do you see AI enhancing ADR measures, and what new quality measures might emerge with AI’s growth?

Dror Zur: In my opinion, AI is nothing less than a game changer in the fight against colorectal cancer, driving improvements in detection rates unmatched by any other technology to date. Independent studies show that every 1% increase in adenoma detection rate (ADR), the proportion of screened patients in whom at least one adenoma is found, reduces the occurrence of colorectal cancer by 3%. That’s why our comprehensive study of our solution, showing a 7% ADR increase, is so exciting.

AI also raises the average number of adenomas found per procedure (APC) and reduces the proportion of adenomas missed (AMR), all without altering the procedure workflow or affecting the patient’s experience. And this is just the tip of the iceberg. At Magentiq Eye, we are advancing AI beyond adenoma detection, pioneering AI based solutions for upper GI detection and disease progress monitoring across the gastrointestinal tract.

Moe: Magentiq Eye has CE approval for polyp type and size estimation. What have you learned from early adoption in Europe, and how do physicians trust these AI-driven insights?

Dror Zur: Our experience is that physicians in Europe see value in real time polyp type and size information as it aids in decision-making, aligning with guidelines like the ESG’s “resect and discard” or “diagnose and leave” strategies. For instance “diagnose and leave” strategy states that endoscopists with a proven accuracy rate on polyp type, are allowed to leave a polyp found in the rectosigmoid in vivo, providing it is hyperplastic and less than 5mm. This combination of real-time characterization and size estimation helps physicians make informed decisions quickly, improving workflow efficiency and patient outcomes. The trust in AI-driven insights grows as physicians see the alignment with established guidelines and the tangible benefits in clinical practice.

Moe: AI in endoscopy is said to reduce physician fatigue and bias. Does Magentiq Eye shift clinician behavior, perhaps changing vigilance, focus, or confidence levels?

Dror Zur: I see our system as a decision support tool, enhancing alertness and reducing missed polyps, however the final decision still remains with the endoscopist. It’s akin to a lane deviation warning system in cars—keeping physicians alert without replacing their expertise. This approach ensures that while AI aids in maintaining high vigilance and reducing cognitive load, it does not replace the clinician’s judgment. The AI serves to augment the physician’s capabilities, allowing them to focus on more complex decision-making tasks while ensuring that routine detections are not overlooked. By maintaining this balance, AI supports clinicians in delivering high-quality care without diminishing their critical role in the decision-making process.

Moe: Magentiq Eye is developing AI tools for various indications. What strategies or frameworks guide your decision on which therapeutic areas to pursue next?

Dror Zur: We focus on areas where AI can significantly impact, such as polyp detection and quality indicators like cecum detection and Boston bowel preparation score. These areas are crucial for procedure quality and align with guidelines. For instance, our new feature of automatic cecum detection ensures that the scope reaches the farthest point in the colon, verifying a complete scan. We also explore specific diseases like IBD, aiming for standardized scoring systems to monitor disease progression and treatment efficacy. By focusing on these areas, we ensure that our AI tools provide tangible benefits in terms of procedure quality and patient outcomes. Our strategy is to prioritize areas where AI can offer the most significant improvements in clinical practice and patient care.

Moe: AI-generated reports could change medical, legal, and reimbursement landscapes. How do you anticipate stakeholders will respond to these machine-derived records?

Dror Zur: I foresee procedure reports becoming largely automatic, with AI covering more features in real-time. This will standardize and streamline report generation, saving time and enhancing accuracy. For example, AI can automatically summarize polyp characteristics and quality indicators, allowing physicians to quickly review and approve reports. This efficiency not only saves time but also ensures more accurate and standardized documentation, which is crucial for patient surveillance and legal compliance. As AI continues to evolve, I expect stakeholders to embrace these advancements for their potential to improve healthcare delivery and reduce administrative burdens. The shift towards AI-generated reports will likely be seen as a positive development, enhancing the quality and efficiency of medical documentation.