Systemic lupus erythematosus (SLE) is a typical systemic autoimmune disease which can affect multiple organ systems, especially kidney. Patients can display severe and progressive kidney damage, namely glomerulonephritis (GN), leading to morbidity and mortality. The search of new medicines against SLE requires pertinent animal model reproducing the pathology and robust readouts for GN evaluation. Until now, in animal lupus models pathologists used to evaluate GN by different scoring systems on a limited number of glomeruli per kidney section. Moreover, the scoring process is time consuming, providing data liable to reproducibility and subjectivity possible issues. The aim of the current study was to propose a digital pathology approach for glomerulonephritis quantification that was able to discard human interpretation while improving the workload.
