Individuals with ulcerative colitis (UC), a chronic inflammatory bowel disease, face a significantly elevated risk of developing colorectal cancer – up to four times higher than the general population. Identifying those at greatest risk, however, has been a clinical challenge. Now, a new advancement leveraging artificial intelligence (AI) is offering a more precise way to predict which patients with low-grade dysplasia (LGD), a potential precursor to cancer, are most likely to see their condition progress.
Researchers at the University of California San Diego have developed an AI workflow that combines clinical data with biostatistical risk models to accurately assess cancer risk in UC patients with LGD. The findings, published in Clinical Gastroenterology and Hepatology, suggest this technology could personalize surveillance intervals and potentially reduce the number of unnecessary colonoscopies.
LGD involves abnormal or precancerous lesions, but crucially, not all cases progress to cancer. This uncertainty often leaves clinicians and patients grappling with difficult decisions regarding continued monitoring versus preventative surgery. The new AI tool aims to provide a clearer picture of individual risk, facilitating more informed choices.
The study involved a large-scale analysis of data from over 55,000 patients within the U.S. Department of Veterans Affairs (VA) health care system. The AI workflow systematically sifted through past medical records, including colonoscopy and pathology reports, to identify UC-LGD patients and evaluate their individual cancer risk. The system demonstrated a particularly strong ability to identify patients at low risk, correctly predicting with approximately 99% certainty that about half of those assessed would remain cancer-free for at least two years.
Beyond identifying low-risk individuals, the AI also revealed that unresectable visible lesions – those that cannot be removed during a colonoscopy – carry a higher cancer risk than previously estimated by clinicians. This finding highlights the importance of accurately characterizing and addressing these lesions.
The AI’s ability to extract and analyze information from clinical notes is a key component of its success. By processing the narrative text within medical records, the system can identify subtle patterns and risk factors that might be overlooked during a traditional review. This automated risk scoring has the potential to streamline the assessment process and provide clinicians with a more comprehensive understanding of each patient’s situation.
The implications of this technology extend beyond simply identifying high-risk patients. By more accurately stratifying risk, clinicians can tailor surveillance schedules to individual needs. Those at low risk may require less frequent colonoscopies, reducing patient burden and healthcare costs. Conversely, those identified as high-risk can be closely monitored or considered for preventative surgery, potentially improving outcomes.
While the study represents a significant step forward, it’s important to note that AI is a tool to aid clinical decision-making, not replace it. The researchers emphasize that the AI workflow is designed to work in conjunction with clinical expertise, providing additional information to inform patient care. Further research will be needed to validate these findings in diverse patient populations and to refine the AI’s predictive capabilities.
The development of this AI tool reflects a growing trend in healthcare towards leveraging the power of large language models and machine learning to improve diagnosis, treatment, and prevention. As AI technology continues to evolve, it holds the promise of transforming the way we approach complex medical challenges, ultimately leading to better care for patients.
The research, , was led by researchers at University of California San Diego and builds on previous work demonstrating the potential of AI in predicting cancer risk. A related study, published around , highlighted the AI’s ability to accurately predict which colitis patients face the highest cancer risk.
