Principal Investigator: Andrea Knight, MD, MSCE
Co-Principal Investigators: Santiago Arciniegas, BSc, Birgit Ertl-Wagner, MD, PhD, MHBA, Matt Wagner, Farzad Khalvati, PhD, MASc, George Ibrahim, MD, PhD, FRCSC, FAANS, Linda Hiraki, MD, FRCPC, ScD, Deborah Levy, MD, MS, FRCPCS.
Study Goals: The goal of this study is to build an algorithm capable of classifying brain MR images of Childhood-onset Systemic Lupus Erythematosus (cSLE) patients.
Description: Retrospective data is collected from de-identified neuroimaging and medical chart records at SickKids. Specifically, we collect multimodal MRI scans and relevant clinical meta-data (e.g., age at time of image acquisition, sex, medical history, etc.) for each participant.
Impact: Conventional MRI reports fail to identify brain abnormalities in over half of NP-cSLE patients. A promising set of modern analytical tools that could be coupled with multimodal neuroimaging includes deep learning (DL) and computer vision methods. These fields have recently seen major growth and success in the automated classification of medical images. We want to incorporate deep learning methods in the study of NP-cSLE to better understand brain changes in patients with cSLE and to build a cSLE classifier.