Abstract / Summary
Perienteric fat stranding on noncontrast CT may reflect appendicitis-associated acute inflammation or Crohn-associated mesenteric phenotype, two conditions that are difficult to distinguish visually. This exploratory study compared quantitative noncontrast CT indices between patients with ileocecal Crohn disease and patients with acute appendicitis, as a reference for appendicitis-associated acute inflammation. This retrospective cross-sectional study included 49 patients with active mild-to-moderate Crohn disease (ileocecal involvement) and 78 with histopathologically confirmed acute appendicitis. Ten quantitative CT indices (perilesional fat density, subcutaneous and mesenteric-root fat density, psoas-muscle attenuation, peritoneal and abdominal height, and four ratios (rCT1, rCT2, rCT3, and the peritoneal-to-abdominal height ratio [PAR])) were measured by two radiologists and averaged. Between-group differences were quantified with logistic regression, leave-one-out cross-validation (LOOCV), and a one-class reference region derived from the appendicitis group alone, and were then examined in a separate cohort of 89 appendicitis cases (59 same-institution, 30 external-center). In a one-class analysis that used only the appendicitis group to define the 95% reference region, only 20.4% (10/49; 95% CI: 10.2, 34.3) of Crohn disease cases fell outside the region, compared with 2.6% (2/78; 95% CI: 0.3, 9.0) of appendicitis cases, indicating substantial overlap between the two conditions. A three-index model (rCT2, PAR, psoas-muscle attenuation) achieved an apparent AUC of 0.880 (95% CI: 0.82, 0.94) and a leave-one-out cross-validated (LOOCV) AUC of 0.854 in this two-group comparison, values that quantify phenotypic separation rather than diagnostic accuracy. Perilesional fat density was lower in Crohn disease than in appendicitis (− 77.4 vs. − 68.1 HU), whereas PAR (0.55 vs. 0.50) and psoas-muscle attenuation (55.1 vs. 50.1 HU) were higher. In an appendicitis-only cohort, 100.0% (59/59) of same-institution (different scanner) and 80.0% (24/30) of external-center cases fell within the 95% reference region, indicating incomplete transportability across scanners and centers; no independent Crohn disease cohort was available. Quantitative noncontrast CT indices differ significantly but incompletely between ileocecal Crohn disease and acute appendicitis. The substantial overlap limits their use as a standalone discriminator; these findings are exploratory and should not be interpreted as a clinically validated diagnostic tool.