Innovation occurs on the fringes of healthcare. Artificial intelligence (AI) has made micro and macro level impacts in healthcare. It has improved operational workflows, diagnostics, and clinical care outcomes, but has only scratched the surface in terms of frontline care implementation. Gastroenterology is being featured as one of the specialties in which AI has potential applicable solutions. AI is a term that defines algorithmic learning from data to create trained outputs by mirroring human thought processes and learning techniques. There are multiple fields of (AI) that are pertinent to frontline healthcare to provide supportive decision-making solutions including computer vision, natural language processing, and augmented reality.

Gastroenterology is enriched with big data that can be utilized with convolutional neural networks to potentially interpret a multitude of studies including small bowel capsule endoscopy, manometry, upper gastrointestinal series, and barium studies as part of the diagnostic testing. There have been multiple studies where images have been labeled for supervised learning and use gradient descent via feed forward neural networks to identify the likely output, in essence, diagnosis in clinical care1

Gastrointestinal bleeding is one of the most common inpatient diagnosis for patients presenting to gastroenterologists. Computed tomography (CT) angiography is a modality used to identify the site of gastrointestinal bleeding. AI can be integrated with CT angiography via utilization of convolutional neural networks and supervised labeled datasets to identify the presence and location of gastrointestinal bleeds on imaging sequences2. Magnetic resonance cholangiopancreatography (MRCP) is also a radiological modality whereby imaging parameter optimization can be controlled by machine learning leading to the most favorable image acquisition for specific disease identification3.

Patients with Irritable bowel syndrome or inflammatory bowel disease have very specific needs. The ability to provide personalized medication regimens can reduce their flare-ups, leading to significant long-term benefits. Using machine learning and gradient descent, patterns in remission can be identified, and relapse episodes due to potential triggers can be recognized4. This can represent a new form of precision medicine with improved treatment regimens based on objective data. 

In interventional gastroenterology, obtaining access to the bile duct is highly critical for the gastroenterologist and is a crucial step of the procedure with risk of complications. With the use of AI, gastroenterologists would be able to utilize real-time images to navigate access to the bile duct through the ampulla for the highest level of success by identifying particular access choices that save time and reduce complications. Utilizing augmented reality to optimize stent size can also prove advantageous for dealing with obstructive jaundice causes during an endoscopic retrograde cholangiopancreatography (ERCP)5

While biliary access is important for interventional gastroenterologists, all gastroenterologists work to remove colonic polyps. With the help of AI utilizing computer vision that has been trained on big data models, polyps may be able to be classified with greater accuracy and patients could be risk stratified on the spot6. This point of care testing could significantly reduce healthcare costs associated with current pathological interpretation of colon polyps. 

Liver transplant registries have matching services, which can automate the process of matching donors with potential recipients. Digitizing and prognosticating the likelihood of rejection will be supported with data models from non-linear time series data. This can also speed up the process and increase the efficiencies before transplant to the correct recipient7.

The deployment of AI in healthcare is a powerful tool. AI technologies are continuing to evolve and become universally implemented. The use of AI will make large volumes of data more manageable and improve clinical efficiency, accuracy, and outcomes across all areas of medicine. AI has potential applications in gastroenterology including improving expertise in interpretation of diagnostic testing, disease identification, and personalized treatment. Its applications can also be seen in interventional gastroenterology to use real-time virtual imaging for more effective treatment and even in liver transplant matching. The convergence of gastroenterology and AI is a trend toward the future of medicine. 


  1. Min, J. K., Kwak, M. S., & Cha, J. M. (2019). Overview of Deep Learning in Gastrointestinal Endoscopy. Gut and Liver, published on 11 January 2019, doi: 10.5009/gnl18384.
  2. Loftus, T. J., Brakenridge, S. C., Croft, C. A., Smith, R. S., Efron, P. A., Moore, F. A., Mohr, A. M., & Jordan, J. R., (2017). Neural network prediction of severe lower intestinal bleeding and the need for surgical intervention. Journal of Surgical Research, 212, 42-47. doi: 10.1016/j.jss.2016.12.032.
  3. Wang, L., Shi, Y., Suk, H. I., Noble, A., Hamarneh, G. (2019). Special issue on machine learning in medical imaging. Computerized Medical Imaging and Graphics, 74, 10-11. doi: 10.1016/j.compmedimag.2019.03.003.
  4. Melidis, C., Denham, S. L., & Hyland, M. E. (2018). A test of the adaptive network explanation of functional disorders using a machine learning analysis of symptoms. Biosystems, 165, 22-30. doi: 10.1016/j.biosystems.2017.12.010.
  5. Jovanobiv, P., Salkic, N. N., Zerem, E. (2014). Artificial neural network predicts the need for therapeutic ERCP in patients with suspected choledocholithiasis. Gastrointestinal Endoscopy, 80(2), 260-268. doi: 10.1016/j.gie.2014.01.023.
  6. Wang, P., Berzin, T. M., Glissen Brown, J. R., et. al. (2019). Real-time automatic detection system increases colonoscopic polyp and adenoma detection rates: a prospective randomised controlled study. Gut, published on 27 February 2019. doi: 10.1136/gutjnl-2018-317500.
  7. Khalaileh, A., Khoury, T., et. al. (2019). Multiplication product of Model of End-stage Liver Disease and Donor Risk Index as predictive models of survival after liver transplantation. European Journal of Gastroenterology & Hepatology, published ahead of print. doi: 10.1097/MEG.0000000000001396.

Love the story? Most guest posts can be read on AIMed blog.

Author Bio

Dr. Sunny Malhotra is a US trained Sports Cardiologist in New York. He is an entrepreneur and health technology investor. He is the founder of, a healthcare focused AI automation company which automates clinical and administrative workflows. He is the winner of the Governor General Caring Canadian Award 2015, NY Superdoctors Rising Stars 2018 & 2019. Twitter: @drsunnymalhotra