Artificial intelligence to aid the detection of anti-citrullinated protein autoantibodies with a hemagglutination assay
Project description:
Rheumatoid arthritis (RA) is an autoimmune disease that affects 18 million people worldwide. It is characterized by chronic inflammation, where the body’s own immune cells are dysregulated and attack healthy tissue. This leads to swelling, pain and stiffness of the joints. Biomarkers that are used for diagnosing patients include anti-citrullinated protein autoantibodies (ACPA). The current golden standard for ACPA-detection relies on enzyme-linked immune assays, which are time consuming, expensive and require highly trained professionals. Here, we develop a faster, cheaper and more easy to use method which relies on hemagglutination, a method that is also frequently applied for blood group typing. A bivalent mediator is generated which on one side recognizes a common and abundant receptor on the surface of red blood cells, and on the other side contains an ACPA-target peptide. When added to a drop of blood from a patient, which is ACPA-positive, this will lead to agglutination of the red blood cells.
Aim:
Develop an artificial intelligence-based tool for the interpretation of hemagglutination data from images and videos to aid ACPA detection in RA patients.
Expected goals:
Train an artificial intelligence (AI) model with images taken during experiments to detect ACPA-dependent hemagglutination. This model will enable faster recognition of the hemagglutination levels and more reliable interpretation of the results.