The following is a summary of “Radiomics to Detect Inflammation and Fibrosis on Magnetic Resonance Enterography in Stricturing Crohn’s Disease,” published in the May 2024 issue of Gastroenterology by Chirra et al.

Magnetic resonance enterography (MRE) is a great tool for diagnosing Crohn’s disease (CD) strictures, but it can’t fully assess inflammation and fibrosis severity within those strictures.

Researchers conducted a retrospective study to develop and validate a machine-learning model that analyzes MRE images to assess inflammation and fibrosis in CD strictures, comparing it to traditional radiologist scoring

They involved 51 patients diagnosed with CD (n=34 for discovery; n=17 for validation) who had confirmed terminal ileal strictures on diagnostic MRE within 15 weeks of resection. Histopathological specimens were examined for inflammation and fibrosis and matched with pre-surgical MRE sequences. Radiologists visually assessed annotated stricture regions on MRE and employed 3D radionics-based machine learning analysis. Histopathology results were compared with both approaches.

The results showed that two distinct sets of radiomic features, capturing textural heterogeneity within strictures, were associated with severe inflammation or severe fibrosis across both discovery (AUC=0.69, 0.83) and validation (AUCs=0.67, 0.78) cohorts. Radiologist visual scoring achieved an AUC=0.67 for severe inflammation and AUC=0.35 for severe fibrosis. The combination of radiomics and radiologist scoring enhanced the identification of severe inflammation (AUC=0.79) and a slightly improved assessment of severe fibrosis (AUC=0.79) compared to individual approaches.

Investigators found that radiomic analysis of MRE images accurately identified severe inflammation and fibrosis in CD strictures, even improving traditional radiologist scoring.

Source: academic.oup.com/ecco-jcc/advance-article-abstract/doi/10.1093/ecco-jcc/jjae073/7676337