META-LUNE: Meta-learning in the Management of Lupus Nephritis

Systemic lupus erythematosus is a serious chronic disease that causes skin lesions and joint pain but can also affect vital organs. Renal involvement is the most common severe manifestation of lupus, often leading to end-stage renal failure requiring hemodialysis and is therefore an important cause of mortality. Renal involvement during lupus is diagnosed by the renal biopsy, which helps to distinguish several different forms of the disease. The treatment of renal lupus is based on immunosuppressive drugs (i.e., drugs that act on the immune system) to control inflammation and decrease the risk of kidney loss. Although several strategies of varying intensities are available, their effectiveness in controlling the disease in any given case is not predictable, with overall success rates of 60-70%. The goal of my project is to develop machine learning tools capable of predicting the response rates of renal lupus to different therapeutic strategies, using the data contained in the renal biopsy. Such tools could help physicians to directly choose the right treatment for patients with renal lupus, increasing their chance of remission and reducing the risk of adverse effects due to drug accumulation.


Academic Year
Area of Study