Fully automated histological classification of cell types and tissue regions of celiac disease is feasible and correlates with the Marsh score

Aims: Histological assessment is essential for the diagnosis and management of celiac disease. Current scoring systems, including modified Marsh (Marsh–Oberhuber) score, lack inter-pathologist agreement. To address this unmet need, we aimed to develop a fully automated, quantitative approach for histology characterization of celiac disease.

Methods: Convolutional neural network models were trained using pathologist annotations of hematoxylin and eosin-stained biopsies of celiac disease mucosa and normal duodenum to identify cells, tissue and artifact regions. Human interpretable features were extracted and the strength of their correlation with Marsh scores were calculated using Spearman rank correlations.

Results: Our model accurately identified cells, tissue regions and artifacts, including distinguishing intraepithelial lymphocytes and differentiating villous epithelium from crypt epithelium. Proportional area measurements representing villous atrophy negatively correlated with Marsh scores (r=−0.79), while measurements indicative of crypt hyperplasia and intraepithelial lymphocytosis positively correlated (r=0.71 and r=0.44, respectively). Furthermore, features distinguishing celiac disease from normal colon were identified.

Conclusions: Our novel model provides an explainable and fully automated approach for histology characterization of celiac disease that correlates with modified Marsh scores, facilitating diagnosis, prognosis, clinical trials and treatment response monitoring.