Artificial intelligence-based comparison of the effects of duodenojejunostomy and sleeve gastrectomy on pancreatic morphology in Zucker diabetic fatty rats

Objective This study aims to compare the effects of sleeve gastrectomy (SG), duodenojejunostomy (DJOS), and their combination (DJOS + SG ) on glucose regulation and pancreatic histomorphology and function in Zucker diabetic fatty (ZDF) rats, using artificial intelligence (AI)–assisted tissue analysis to assess morphological alterations. Methods Forty-five male ZDF rats were randomized into three surgical groups (SG, DJOS, DJOS + SG ). Oral glucose tolerance tests (OGTT) and insulin levels were assessed at 1, 3 and 6 months post-surgery. Pancreatic tissue was analyzed histologically and immunohistochemically for β -cell mass, PCNA and PDX-1 expression. QuPath software enabled AI-based quantification of acinar, adipose, and fibrotic tissue. Results DJOS and DJOS + SG improved glucose tolerance and increased fasting insulin compared to SG. Both bypass groups demonstrated greater β -cell mass and clustering, elevated PCNA and PDX-1 expression, and more acinar tissue. SG was associated with reduced β -cell presence and increased pancreatic adiposity. Conclusion Malabsorptive (DJOS) or combination bariatric procedures (DJOS + SG ) significantly enhance glycemic control in the rat model. These effects are accompanied by increased β -cell numbers and clustering, as well as enhanced β -cell proliferation and differentiation. Furthermore, acinar glandular tissue is increased, while pancreatic adiposity is reduced following bypass surgery.