Objective IgA nephropathy (IgAN) and IgA vasculitis (IgAV) are part of a similar clinical spectrum. Both clinical conditions occur with the coronavirus disease 2019 (COVID-19). This review aims to recognize the novel association of IgAN and IgAV with COVID-19 and describe its underlying pathogenesis. Methods We conducted a systematic literature search and data extraction from PubMed, Cochrane, ScienceDirect, and Google Scholar following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Results Our search identified 13 cases reporting IgAV and IgAN associated with COVID-19 infection and 4 cases of IgAN following COVID-19 vaccination. The mean, mode, and median ages of patients were 23.8, 4, and 8 years, respectively. Most cases associated with COVID-19 infection were reported in males (76.9%). Rash and purpura (84.6%) were the most common clinical features, followed by gastrointestinal symptoms (61.5%). In symptomatic cases, skin or renal biopsy and immunofluorescence confirmed the diagnosis of IgAN or IgAV. Most patients were treated with steroids and reported recovery or improvement; however, death was reported in two patients. Conclusion There is a paucity of scientific evidence on the pathogenesis of the association of IgAN and IgAV vasculitis with COVID-19, which thus needs further study. Current research suggests the role of IgA-mediated immune response, evidenced by early seroconversion to IgA in COVID-19 patients and the role of IgA in immune hyperactivation as the predominant mediator of the disease process. Clinicians, especially nephrologists and paediatricians, need to recognize this association, as this disease is usually self-limited and can lead to complete recovery if prompt diagnosis and treatment are provided.

No abstract

We propose a tensor network that can learn to perform multiple tasks by adjusting the factors of each layer. Most of the existing methods for multi-task learning train a single network to extract task-specific features and subsequent prediction. We propose to use a single network with task-specific transformations that can extract task-specific features and perform task inference with small memory overhead. In particular, we transform features using low-rank updates in the convolution kernels. We present experiments on different datasets for multi-task and multi-domain learning and demonstrate that our method achieves state-of-the-art performance with minimal memory overhead compared to existing methods.