Mod Pathol. 2023 Jul 18:100285. doi: 10.1016/j.modpat.2023.100285. Online ahead of print.
ABSTRACT
We have developed an artificial intelligence (AI)-based digital pathology model for the evaluation of histologic features related to eosinophilic esophagitis (EoE). In this study, we evaluate the performance of our AI model in a cohort of pediatric and adult patients for histologic features included in the Eosinophilic Esophagitis Histologic Scoring System (EoEHSS). We collected a total of 203 esophageal biopsies from patients with mucosal eosinophilia of any degree (91 adults, 112 pediatric patients) and 10 normal controls from a prospectively maintained database. All cases were assessed by a specialized gastrointestinal (GI) pathologist for features in the EoEHSS at the time of original diagnosis and re-scored by a central GI pathologist. We have subsequently analyzed whole-slide image digital slides using a supervised AI model operating in a cloud-based, deep-learning artificial intelligence platform (Aiforia Technologies, Helsinki, Finland) for peak eosinophil count (PEC) and several histopathologic features in the EoEHSS. The correlation and inter-observer agreement between the AI model and pathologists (rs=0.89 and ICC=0.87 vs. original pathologist [OP]; rs =0.91 and ICC=0.83 vs. central pathologist [CP]) was similar to the correlation and inter-observer agreement between pathologists for PEC (rs=0.88 and ICC=0.91) and broadly similar for most other histologic features in the EoEHSS. The AI model PEC also accurately identified >15 eosinophils/HPF by the OP (AUC [area under the curve]=0.98) and CP (AUC=0.98) and had similar AUCs for the presence of EoE-related endoscopic features than pathologists’ assessment. Average eosinophils per epithelial unit area had similar performance compared to AI HPF-based analysis. Our newly developed AI model can accurately identify, quantify, and score several of the main histopathologic features in the EoE spectrum, with agreement regarding EoEHSS scoring, which was similar to that seen among GI pathologists.
PMID:37474003 | DOI:10.1016/j.modpat.2023.100285
