2026 Update from a UK-trained gastroenterologist in Dubai.
Artificial intelligence is changing many parts of modern life. One area where I think it may genuinely improve medical care is inflammatory bowel disease, especially Crohn’s disease and ulcerative colitis.
In IBD, one of the hardest questions is simple: can we predict which treatment is most likely to work before losing months on trial and error?
In this video, I explain a new Crohn’s disease study where machine learning was used to analyse biological signals from blood, stool, metabolism and the gut microbiome before treatment started.
You can watch the video below. If you prefer reading, the key points are explained underneath.
Short Answer
AI may help in Crohn’s disease and ulcerative colitis by analysing biological signals from blood, stool, metabolism and the gut microbiome to predict who is more likely to respond to treatment.
In one Crohn’s disease study, a machine learning model predicted response to exclusive enteral nutrition better than usual clinical and colonoscopy based assessment.
This is not routine clinic testing yet, but it shows how IBD care may move toward more precise treatment prediction.
How Could AI Help Predict Treatment Response in Crohn’s and Ulcerative Colitis?
| Question patients ask | What doctors usually use now | What AI and machine learning may add | Why this matters |
|---|---|---|---|
| Can we predict if a Crohn’s or colitis treatment will work? | Symptoms, blood tests, stool calprotectin, colonoscopy, scans, disease location and previous treatment history. | AI may analyse many biological signals together, including blood, stool, metabolism, lipid patterns and the gut microbiome. | This may help identify likely responders earlier, before waiting weeks or months to see if treatment works. |
| What did the Crohn’s disease study show? | Standard clinical and colonoscopy based assessment was compared with deeper biological data. | The machine learning model predicted response to exclusive enteral nutrition better than usual clinical and colonoscopy based assessment. | This suggests AI may detect biological patterns that routine assessment does not fully capture. |
| Was this about biologics? | No. The study used exclusive enteral nutrition, a diet based therapy used especially in children with Crohn’s disease. | The same prediction principle may be applied more and more to biologics, small molecules and treatment sequencing in IBD. | This could make future treatment choices more precise and less dependent on trial and error. |
| Is AI ready to choose IBD treatment today? | No. Treatment decisions still need a gastroenterologist and full clinical review. | AI may become a support tool to help interpret complex biological patterns. | The future is not replacing doctors. It is helping doctors make better, earlier and more personalised decisions. |
What Was the Main Question in the Crohn’s Disease AI Study?
In this particular study, the treatment was exclusive enteral nutrition, often shortened to EEN. This is a liquid nutrition based treatment used in Crohn’s disease, especially in children and teenagers.
The researchers were not asking whether AI could replace doctors. They were asking whether machine learning could analyse biological information in a deeper way and predict treatment response before the outcome became obvious clinically.
What Is Exclusive Enteral Nutrition?
Exclusive enteral nutrition means receiving all nutrition from a prescribed liquid formula for a defined period, with no regular food during that phase.
It is used particularly in pediatric Crohn’s disease and can be very effective for inducing remission. In simple terms, it gives the bowel a controlled nutritional environment while still providing calories, protein, vitamins and minerals.
What Did the Crohn’s Disease AI Study Actually Do?
The researchers studied children and teenagers with newly diagnosed Crohn’s disease who were about to start exclusive enteral nutrition.
Before treatment, they analysed several layers of biological information. These included blood signals, stool signals, lipid patterns, metabolism and the gut microbiome.
This approach is called multi omics.
Instead of looking at only one blood test, one stool marker or one colonoscopy finding, multi omics combines several biological layers together. Machine learning then looks for patterns across these layers.
The question put to the model was simple:
Based on these biological signals, who is more likely to respond to exclusive enteral nutrition?
What Did the AI Model Find?
The machine learning model predicted response to exclusive enteral nutrition better than usual clinical and colonoscopy based assessment.
That is the part I find remarkable.
As gastroenterologists, we already look at symptoms, blood tests, stool markers, scans, colonoscopy findings, ulcer severity, disease location and previous treatment history. These are important and will remain important.
But in this study, the AI model appeared to detect a deeper biological pattern that routine clinical assessment was not fully capturing.
That does not mean doctors are poor at making decisions.
It means IBD biology is complex, and some patterns may be too subtle for the human eye alone.
What Does Multi Omics Mean in Simple Language?
Multi omics means looking at several biological layers together.
| Biological layer | Simple meaning | Why it may matter in IBD |
|---|---|---|
| Gut microbiome | The bacteria and microbes living in the gut. | May influence inflammation, treatment response and gut barrier function. |
| Metabolomics | Small chemical signals produced by the body, diet and gut microbes. | May show how the gut environment is behaving during inflammation. |
| Lipidomics | Fat related molecules involved in cell signalling and inflammation. | May reveal inflammatory pathways not seen on routine tests. |
| Blood and stool signals | Markers that reflect immune activity and gut inflammation. | May help connect symptoms, inflammation and deeper biology. |
Why Is This Relevant to Biologics in Crohn’s and Ulcerative Colitis?
This particular study was about diet therapy in Crohn’s disease, not biologic treatment.
But the principle is important.
If AI can use biological signals to predict response to one treatment, the same idea may increasingly be applied to biologics, small molecules and advanced treatment sequencing in Crohn’s disease and ulcerative colitis.
That matters because IBD treatment is still too often affected by uncertainty. A person may start a biologic and only know weeks or months later whether it is working well enough.
In the future, the goal is not just to ask which treatment should we try next. The better question is which treatment is most likely to match the biological pattern driving this person’s inflammation.
Can AI Already Choose the Best Biologic for IBD?
No. AI should not currently be used on its own to choose a biologic for Crohn’s disease or ulcerative colitis.
Treatment decisions still require a qualified gastroenterologist who understands the full clinical picture, including disease severity, disease location, previous treatment exposure, safety risks, blood tests, stool markers, scans, colonoscopy findings and patient preference.
But research is moving toward better prediction.
Multi omics studies are already exploring whether biological signatures can help predict response to biologics and small molecule therapies. This is where I think the field is heading.
Is This the Same as Asking ChatGPT for Medical Advice?
No.
This is a very important distinction.
The study discussed here was not about asking a public chatbot what treatment to take. It used machine learning to analyse biological data.
Public AI tools may be useful for education, summarising information or helping people understand medical concepts. But they should not be used to choose personal treatment without proper medical review.
The most exciting medical use of AI in IBD may be deeper than general advice: helping clinicians identify biological patterns that are too complex to interpret manually.
Why Does Predicting Treatment Response Matter to Patients?
For someone with Crohn’s disease or ulcerative colitis, the most frustrating part of treatment is uncertainty.
You may start a treatment and wait weeks or months to know whether it is working. If it fails, inflammation may continue during that time. That can mean symptoms, steroid exposure, nutritional problems, complications and emotional stress.
Better prediction could change that.
If future AI models can help identify which treatment is more likely to work for which person, IBD care could become more precise and less dependent on trial and error.
This does not mean AI will replace gastroenterologists.
It means AI may help gastroenterologists make better decisions, earlier.
Why Did I Find This Paper Exciting?
I found this paper exciting because it shows the direction of travel.
For years, IBD treatment has improved mainly through better medicines. We now have more biologics, more small molecules and better treatment strategies than before.
But the next major shift may be prediction.
Not just asking which treatment do we have, but asking which treatment is most likely to work for this person.
That is a very different question.
This study was about diet therapy in pediatric Crohn’s disease. But the same thinking is now moving toward biologics, treatment response prediction and more personalised IBD care.
I am seeing this change in real time.
What Should Patients Take Away From This?
The main message is not that AI is ready to choose your Crohn’s or colitis treatment today.
The message is that IBD care is moving toward a more precise future.
AI and machine learning may help us understand which biological patterns predict treatment response. Multi omics may help us look beyond symptoms and colonoscopy alone. Over the next few years, this could become increasingly important in how we choose, monitor and adjust treatment in Crohn’s disease and ulcerative colitis.
This is not routine clinic testing yet.
But I genuinely think we are on the cusp of major changes in IBD care.
FAQ
Can AI currently choose my Crohn’s disease treatment?
No. AI should not be used on its own to choose Crohn’s disease treatment. Treatment decisions still need a qualified gastroenterologist who understands your full clinical history, test results, disease severity, safety risks and treatment goals.
Was this study about biologics?
No. The study discussed in this article was about predicting response to exclusive enteral nutrition in pediatric Crohn’s disease. However, the principle of using AI and multi omics to predict response may become increasingly relevant to biologics and advanced IBD therapies.
What is exclusive enteral nutrition?
Exclusive enteral nutrition is a treatment where all nutrition comes from a prescribed liquid formula for a set period. It is used especially in children and teenagers with Crohn’s disease and should be supervised by a specialist team.
What does AI look at in this type of research?
In this type of research, AI may analyse biological signals from blood, stool, metabolism, lipid patterns and the gut microbiome. The aim is to find patterns that may predict treatment response.
Does AI replace colonoscopy in IBD?
No. Colonoscopy remains important in Crohn’s disease and ulcerative colitis. AI may become an additional tool to help interpret complex biological data, but it does not replace proper assessment, monitoring or specialist review.
Why is predicting treatment response important in IBD?
Predicting treatment response matters because Crohn’s disease and ulcerative colitis treatments can take time to work. If we can identify earlier who is more likely to respond, we may reduce unnecessary delays, ongoing inflammation and avoidable treatment failure.
Could AI help predict biologic response in the future?
Yes. That is one of the most important areas of future research. Multi omics and machine learning models are already being explored to predict response to biologics and other advanced therapies in IBD.
Can patients use public AI tools to decide treatment?
No. Public AI tools may help explain general medical concepts, but they should not be used to decide personal Crohn’s disease or ulcerative colitis treatment. IBD treatment decisions need proper specialist review.
Recommended Guides: AI, Omics and Treatment Prediction
Useful background on how omics research may help predict biologic treatment response in IBD.
Explains why treatment response varies and why prediction matters in modern IBD care.
Discusses biomarker based prediction and future treatment selection in IBD.
Looks at the future of predicting response to anti TNF treatment in Crohn’s and colitis.
Explains how early biomarker changes may help guide vedolizumab treatment decisions.
Another example of future testing that may refine inflammation monitoring in Crohn’s and colitis.
Biologics and Advanced IBD Treatments
Learn about Biologicals used in UC and Crohn’s in Dubai.
A newer oral therapy concept targeting a pathway already important in modern IBD care.
Relevant to future treatment sequencing and pathway based therapy selection.
Explains where small molecule therapy may fit in advanced IBD treatment.
Diet, Enteral Nutrition, Microbiome and Response
Relevant for readers interested in nutrition, Crohn’s disease and biologic response.
Explores how diet and the microbiome may influence biologic treatment response.
Explains how diet, microbiome and inflammation interact in Crohn’s disease and ulcerative colitis.
Useful for patients interested in prediction, calprotectin, diet and flare risk.
A diet based research topic showing how the gut environment may influence inflammation.
Relevant to microbiome, immune response and supportive IBD care.
Microbiome Therapies and Future IBD Prediction
Discusses microbiome based strategies being explored alongside biologic therapy.
Explains why microbiome based therapy remains an important research area in IBD.
Shows how defined microbial therapies are being studied in Crohn’s and ulcerative colitis.
Useful for readers interested in microbiome targeted treatment and future precision IBD care.
Highlights another microbiome therapy being studied in ulcerative colitis.
Relevant for readers interested in microbiome signals and future prediction tools.
Related Crohn’s and Colitis Pages
Overview of Crohn’s disease symptoms, tests, monitoring and modern treatment options.
Overview of ulcerative colitis symptoms, flares, monitoring and treatment options.
More patient friendly updates on Crohn’s disease, ulcerative colitis, biologics, microbiome and precision IBD care.
If you wish to discuss this further or would like to make an appointment, please use the booking form on this page.