Artificial intelligence in gastroenterology and hepatology

GI disorders are among the leading causes of healthcare burden in the United States and patients with GI disorders generate an abundance of health data, including documented history and physical examination, laboratory tests, radiologic studies, endoscopic images, and tissue samples. The emergence of AI presents exciting opportunities to utilize this data to advance the care of patients with GI disorders. Advanced Machine Learning (ML) methods can yield novel insights on disease risk factors and phenotypes and Deep Learning (DL) algorithms can rapidly and accurately process unstructured, high-dimensional data, such as texts, images, and waveforms. AI is poised to revolutionize the field of diagnostic endoscopy in the near future. This chapter provides an overview and highlights some notable examples of the many potential applications of AI in the field of GI. The examples given for AI applications in this field encompass diseases of the esophagus (gastroesophageal reflux disease, esophageal motility disorders, Barrett’s esophagus, and esophageal adenocarcinoma), the stomach (peptic ulcer disease and gastric cancer), the small intestine, the colon (colon polyp detection and inflammatory bowel disease (IBD)), the liver (fibrosis, steatosis, lesions, and cancers), and the pancreas (pancreatitis, pancreatic cancer). Lastly, the current limitations within this field and more generally to the implementation of AI for patient care are outlined.