What is it about?
Inflammatory bowel disease (IBD), including Crohn’s disease and ulcerative colitis, is a lifelong condition that affects millions of people. Although many new treatments are available, it can still be difficult to understand why some patients respond to treatment while others do not or experience side effects. One reason is that important information is spread throughout the medical record, including laboratory results, diagnoses, imaging, pathology reports, and physicians’ clinical notes. We developed an IBD Data Lake that brings these different sources of information together and uses artificial intelligence and natural language processing to help find and organize relevant clinical information. We tested the system against manual review of medical records and found that it could accurately identify patients with IBD and extract important clinical characteristics. This approach could make it faster and easier for clinicians and researchers to study IBD and identify specific groups of patients for future research.
Featured Image
Photo by julien Tromeur on Unsplash
Why is it important?
High-quality clinical data are essential for improving IBD care and answering important research questions, but collecting information manually from medical records is time-consuming and difficult to maintain. Our study shows that artificial intelligence can help overcome this challenge by bringing together information from different parts of the medical record and rapidly extracting clinically relevant information from physicians’ notes. The IBD Data Lake accurately identified patients with IBD and several important disease characteristics, demonstrating its potential to support research without requiring extensive manual data collection. By making real-world clinical data easier to access and organize, this approach could help researchers study specific patient populations more efficiently and ultimately support more personalized IBD care. The infrastructure also provides a foundation for incorporating newer technologies, including large language models and generative AI.
Perspectives
I find it exciting that we can use artificial intelligence to turn the large amount of information already contained in our patients’ medical records into something that is much easier to search and study. What started as an effort to reduce the time spent manually collecting data has the potential to become a much broader platform for IBD research. I hope this work encourages clinicians and researchers to think differently about how we use clinical data and opens the door to new ways of studying IBD, including using emerging technologies such as large language models and generative AI. Ultimately, the goal is not simply to collect more data, but to use the information we already have to better understand our patients and improve their care.
Dr. Jeremy Liu
University or Montreal Hospital Centre
Read the Original
This page is a summary of: A novel inflammatory bowel disease registry powered by artificial intelligence and natural language processing, PLOS Digital Health, August 2026, PLOS,
DOI: 10.1371/journal.pdig.0001603.
You can read the full text:
Contributors
The following have contributed to this page
