Explore chapters and articles related to this topic
Lupus Nephritis
Published in Richard K. Burt, Alberto M. Marmont, Stem Cell Therapy for Autoimmune Disease, 2019
Less commonly, some patients present with renal disease without any systemic symptoms or signs of SLE. In such cases, renal biopsy may reveal pathologic changes consistent with lupus nephritis. For instance, SLE with membranous nephritis can be distinguished from idiopathic membranous nephropathy by a biopsy showing a typical pattern of immune deposition (mesangial and subendothelial as well as subepithelial deposits). Also, deposition of all isotypes of immunoglobulin is characteristic of SLE, as opposed to idiopathic membranous nephritis where only IgG is commonly present. These patients may exhibit symptoms and serologic markers for SLE only years after the diagnosis of renal lupus.17
Uncertainty-aware membranous nephropathy classification: A Monte-Carlo dropout approach to detect how certain is the model
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2023
Paulo Chagas, Luiz Souza, Izabelle Pontes, Rodrigo Calumby, Michele Angelo, Angelo Duarte, Washington Lc-Dos Santos, Luciano Oliveira
Membranous nephropathy (MN) is a common autoimmune glomerular disease, usually associated with the cause of nephrotic syndrome in adults. The main characteristic of MN is the large quantity of immune complex sediments on the epithelial cells, visually indicated by a thickening in the glomerular basement membrane. Figure 1 depicts a normal glomerulus (left) and another one with MN (right), where we can identify the presence of thickened membranes (boundaries of white areas inside the right glomerulus). Distinguishing these visual characteristics is not a trivial task, demanding trained pathologists that do not always reach a consensus. This way, automatic classification approaches can aid pathologists in the decision-making pipeline. By developing deep learning models, computer vision applications have significantly advanced through time and its full potential for diagnostic-driven studies is still being investigated (Lit- jens et al. 2017).