Determination of the Relative Frequencies of Expected Diagnoses in Duodenal Biopsies: An Essential Step in Developing an Artificial Intelligence Approach to Diagnostic Classification

Understanding the diagnostic landscape prior to developing novel diagnostic strategies is key to managing expectations and authenticating results. In considering the possibility of developing alternate diagnostic approaches for coeliac disease based on duodenal biopsies, we audited 18 months’ worth of duodenal biopsies received in our centre to determine the exact proportions of different diagnoses. A total of 6245 duodenal biopsies were audited, of which 73.76% were normal and 8.84% fell within the spectrum of coeliac disease. Additionally, 6.47% were classified as showing non-specific inflammation, 1.86% were adenomas, 0.42% were carcinomas, and 0.06% were neuroendocrine tumours. Rarer diagnoses included ulceration, Helicobacter pylori infection, Giardiasis, lymphangiectasia, transplant rejection, and lymphoma. Furthermore, 227 biopsies (3.63%) showed isolated intraepithelial lymphocytosis, of which 24 cases eventually received a definitive diagnosis of coeliac disease. We present the first long-term audit of all endoscopic duodenal biopsies received by the histopathology department of a tertiary care facility. The results indicate that a fully automated system that could identify normal duodenal biopsies and biopsies within the spectrum of coeliac disease-associated enteropathy could decrease pathologists’ endoscopic duodenal biopsy workload by up to 80%.